Lower gastrointestinal bleeding (LGIB), bleeding that arises from the colon, rectum, or anus, is among the more frequent problems encountered in gastroenterology and contributes substantially to emergency admissions to gastroenterology units. Its severity varies widely, from minor anorectal bleeding that stops spontaneously to massive, life-threatening hemorrhage that demands prompt resuscitation, inpatient care, endoscopy, or operative intervention. The diversity of underlying causes and the difficulty in predicting severity when patients first present make LGIB an important clinical challenge [1,2,3].

The etiology of LGIB varies widely depending on patient age, comorbidities, and regional disease patterns. Causes include benign anorectal diseases, such as hemorrhoids and anal fissures; inflammatory bowel disease; infectious and nonspecific colitis; diverticular disease; vascular lesions, such as angiodysplasia; and colorectal neoplasia. This heterogeneity frequently complicates first-line clinical assessment and necessitates careful early diagnostic evaluation for appropriate management. In particular, the ability to differentiate benign, self-limiting conditions from clinically significant disease is crucial to minimizing morbidity and avoiding missed diagnoses of serious disease [4,5].

Colonoscopy is considered the gold standard for the diagnosis of LGIB, as it allows direct visualization of the colonic mucosa, localization of the bleeding source, and simultaneous diagnostic and therapeutic intervention. Its applications include endoscopic tattooing of lesions for surgical planning, polypectomy, endoscopic hemostasis, and biopsy of suspicious lesions. Advances in endoscopic techniques have greatly facilitated the management of LGIB, reducing the need for emergency surgery and improving patient outcomes in well-resourced health care systems [6,7].

In resource-limited settings, however, the diagnosis and management of LGIB can be challenging. Limited access to emergency colonoscopy services, insufficient training of endoscopists, inadequate endoscopic equipment, and delayed patient presentation contribute to suboptimal diagnostic timing. These constraints can result in more severe disease at presentation, longer hospitalization, and greater requirements for blood transfusion and inpatient care. Additionally, differences in health care infrastructure can affect clinical decision-making and may cause referrals for endoscopy to depend on disease severity rather than on clinical indication alone [8,9].

Despite these limitations, endoscopic evaluation should be considered the cornerstone of LGIB management in low-resource settings. Understanding real-world clinical outcomes in these contexts is crucial for improving service provision, resource allocation, and evidence-based protocols tailored to local health systems. Assessment of the spectrum of colonoscopic findings is also informative regarding regional disease patterns, which may differ from those reported in high-income countries [10].

The objective of this study was to assess the colonoscopic findings, therapeutic interventions, and clinical outcomes of patients who presented with LGIB in a resource-limited health care setting. This study also examined the associations of demographic factors and laboratory parameters (hemoglobin) with key endpoints, such as hospital admission and transfusion requirements, with the aim of identifying clinically relevant predictors of disease severity and health care utilization.

We conducted this retrospective, observational, analytical study over 2 years (January 2024 to December 2025) at the Combined Military Hospital, Peshawar, Pakistan, a tertiary care center operating with limited resources. Our purpose was to examine outcomes in patients with LGIB and to define the diagnostic and therapeutic contributions of colonoscopy in everyday practice. The analysis relied on real-world data that mirrored routine workflows, endoscopy availability, and the resource constraints of the center. The study cohort consisted of 153 patients whose clinical features and findings were LGIB. We reviewed consecutive patients referred for colonoscopy because of rectal bleeding, a change in bowel habits accompanied by bleeding, or anemia suspected to originate from the lower gastrointestinal tract. Only patients aged 18 years or older were eligible. Eligibility required a clinical picture compatible with LGIB, completion of a diagnostic colonoscopy, and the availability of complete clinical and laboratory records. We excluded patients whose bleeding was traced to an upper gastrointestinal source, those whose colonoscopy was incomplete or nondiagnostic owing to inadequate bowel preparation or technical difficulty, and those whose key clinical or laboratory data were missing. Information was extracted from three sources: hospital electronic medical records, endoscopy reports, and the laboratory information system. The variables recorded included demographic characteristics (age and sex); clinical features (symptom duration and hemodynamic status); laboratory results (hemoglobin); colonoscopic diagnoses (hemorrhoids, colitis, polyps, malignancy, diverticulosis, angiodysplasia, solitary rectal ulcer, nonspecific colitis, stercoral ulcer, or normal findings); endoscopic procedures (biopsy, polypectomy, tattooing, and electrocautery hemostasis); and clinical outcomes (hospital admission and need for blood transfusion). Colonoscopy was performed in all included patients using a standard video colonoscope under conscious sedation, in accordance with institutional protocols. Bowel preparation was achieved using a standard polyethylene glycol (PEG)-based solution prior to the procedure. A complete colonoscopic examination was systematically performed from the rectum to the cecum, with intubation of the terminal ileum whenever feasible. Endoscopic findings were carefully documented and classified according to their macroscopic appearance. Therapeutic interventions, including biopsy for histopathology, polypectomy for polyp removal, endoscopic tattooing to mark lesion sites, and electrocautery for the control of active bleeding, were carried out as indicated. The type of endoscopic management was tailored to the underlying lesion identified by colonoscopy. Biopsies were taken from inflammatory, suspicious, or neoplastic lesions. Colonic polyps were treated by polypectomy and then classified histopathologically. In selected cases, tattooing was performed to aid surgical planning, especially for malignant or large lesions. Active bleeding requiring immediate control was treated with electrocautery hemostasis. The primary outcomes were hospital admission and blood transfusion requirements. Secondary outcomes included the distribution of colonoscopic diagnoses, the frequency of endoscopic interventions, and the associations of diagnosis, hemoglobin level, and age with clinical outcomes. Statistical analysis was carried out with SPSS version 27.0 and Python version 3.10 (statsmodels package). Depending on their distribution, continuous data are summarized as the mean ± standard deviation (SD) or median (range), and categorical data are summarized as counts with percentages. Age, symptom duration, and hemoglobin concentration were checked for normality with the Kolmogorov–Smirnov and Shapiro–Wilk tests. Continuous variables were compared across diagnostic groups by ANOVA, and pairwise differences were examined with Tukey's post hoc test. Relationships among colonoscopic diagnosis, hospital admission, and transfusion requirements were tested with the chi-square test; because many cells had expected counts below 5, Monte Carlo simulation with 100,000 resamples was additionally used to obtain exact p values. Factors independently associated with hospital admission were identified by binary logistic regression and are expressed as odds ratios (ORs) with 95% confidence intervals (CIs). Statistical significance was set at p < 0.05. Ethical approval was granted by the Institutional Review Board of the Combined Military Hospital, Peshawar, Pakistan (No. 298/2026), and the research adhered to the Declaration of Helsinki. Patient records were deidentified before analysis to safeguard confidentiality. Because only existing records were reviewed retrospectively, the board waived the need for individual informed consent.

3.1. Patient characteristics

The analysis included 153 patients who underwent colonoscopy for LGIB. Men slightly outnumbered women (n = 82; 53.6%), and the mean age was 44.89 ± 14.23 years (median 44 years; range 18–80). Age followed an approximately normal distribution (Kolmogorov–Smirnov, p = 0.200; Shapiro–Wilk, p = 0.116). Symptoms were present for a mean of 17.31 ± 22.58 days before colonoscopy (median 11 days; range 1–139 days). This variable was markedly right skewed (skewness = 2.97; kurtosis = 10.75) and deviated significantly from normality (Kolmogorov–Smirnov, p < 0.001; Shapiro–Wilk, p < 0.001), reflecting the small number of patients who presented after a long symptomatic period.

The mean hemoglobin concentration was 11.79 ± 1.87 g/dL (median 11.9; range 6.3–15.4). The majority of patients presented in a hemodynamically stable state (n = 146; 95.4%), and 7 patients (4.6%) experienced hemodynamic instability. Blood transfusions were needed in 11 patients (7.2%). Hospital admission was necessary for 53 patients (34.6%), while 100 patients (65.4%) were managed on an outpatient basis. Baseline demographic and clinical characteristics are summarized in Table 1.

Table 1. Baseline Demographic and Clinical Characteristics of the Study Population (N = 153).

SD = standard deviation.

3.2. Colonoscopy Findings

Colonoscopy revealed a spectrum of pathological diagnoses across all 153 patients. Hemorrhoids constituted the most frequent finding (n = 70; 45.8%), followed by colitis and colonic polyps (n = 15 each; 9.8% each). Malignant growth and solitary rectal ulcer syndrome were identified in 10 patients (6.5% each). Diverticulosis was found in 9 patients (5.9%), and angiodysplasia was found in 8 patients (5.2%). Nonspecific colitis was documented in 7 patients (4.6%), and a normal colonoscopic examination was recorded in 6 patients (3.9%). The least frequent diagnosis was a stercoral ulcer, affecting 3 patients (2.0%). A comprehensive summary of colonoscopic diagnoses, together with their corresponding rates of hospital admission and transfusion, is presented in Table 2.

Table 2. Colonoscopic diagnoses and associated hospital admission and blood transfusion rates (N = 153).

Note. The percentages in the n (%) column are those of the total cohort; the admission and transfusion percentages are those of patients within each diagnostic group. Association of diagnosis with hospital admission, χ²(9) = 50.285, and with blood transfusion, χ²(9) = 52.305; Monte Carlo exact p < 0.001 for both.

3.3. Endoscopic Interventions

Endoscopic interventions were performed in a subset of patients on the basis of colonoscopic findings. Tissue biopsy was the most frequently performed procedure and was performed in 53 patients (34.6%). Polypectomy was performed in 10 patients (6.5%), predominantly those with colonic polyps confirmed by colonoscopy, with histopathological analysis subsequently revealing hyperplastic polyps, tubular adenomas, and tubulovillous adenomas across this subgroup. Endoscopic tattooing for lesion localization was carried out in 5 patients (3.3%), primarily in the context of polyps and malignant lesions requiring surgical planning. Electrocautery hemostasis was applied in 2 patients (1.3%), both of whom presented with actively bleeding lesions at the time of colonoscopy.

3.4. Comparative Analyses: Age and hemoglobin across diagnostic groups

The mean age differed significantly among the colonoscopic diagnostic groups according to one-way ANOVA (F[9, 143] = 10.063, p < 0.001). Tukey's honestly significant difference (HSD) test revealed that patients with malignant growth (62.8 ± 5.9 years) and those with angiodysplasia (61.9 ± 8.1 years) were older than patients with colitis (43.4 ± 15.4 years), hemorrhoids (39.0 ± 11.6 years), nonspecific colitis (38.7 ± 18.4 years), a normal colonoscopy (37.0 ± 8.8 years), or a solitary rectal ulcer (38.2 ± 9.5 years; p < 0.05 for every pairwise comparison). Patients with diverticulosis (56.6 ± 7.4 years) and polyps (55.3 ± 11.1 years) were likewise older than those with hemorrhoids, a normal colonoscopy, or a solitary rectal ulcer (p < 0.05 for each). There was no significant difference in age between the stercoral ulcer group (49.3 ± 5.1 years; n = 3) and any other group, and the younger groups (those with hemorrhoids, solitary rectal ulcers, nonspecific colitis, and normal findings) did not differ significantly from one another.

The hemoglobin concentration also varied significantly by diagnosis (F[9, 143] = 7.966, p < 0.001). According to the results of Tukey’s test, compared with the angiodysplasia (12.49 ± 1.67 g/dL), diverticulosis (11.86 ± 1.02 g/dL), polyp (12.36 ± 1.36 g/dL), hemorrhoid (12.29 ± 1.49 g/dL), nonspecific colitis (12.80 ± 1.17 g/dL), and normal colonoscopy (13.43 ± 1.94 g/dL) groups (p < 0.05 for all). The hemoglobin concentration in the colitis group (10.13 ± 1.15 g/dL) was likewise significantly lower than that in the angiodysplasia, polyp, hemorrhoid, nonspecific colitis, and normal colonoscopy groups (p < 0.05 for each), and the normal colonoscopy group had significantly higher values than the solitary rectal ulcer group did (p = 0.021). The group means, standard deviations, and post hoc groupings are shown in Table 3.

Table 3. Age and hemoglobin levels by colonoscopic diagnosis.

Note. One-way ANOVA: age, F(9, 143) = 10.063, p < 0.001; hemoglobin, F(9, 143) = 7.966, p < 0.001. Means in the same column with a common superscript letter are not significantly different (Tukey’s HSD, α = 0.05). ANOVA = analysis of variance; HSD = honestly significant difference; SD = standard deviation.

3.5. Association Analyses: Diagnosis with Hospital Admission and Transfusion

Chi-square analysis demonstrated a highly significant association between colonoscopic diagnosis and hospital admission (χ²[9] = 50.285; Monte Carlo exact p < 0.001). All patients with malignant growth (10/10; 100%) and stercoral ulcers (3/3; 100%) required hospital admission. Colitis had the next highest admission rate, at 80.0% (12/15), followed by solitary rectal ulcers, at 40.0% (4/10). In contrast, hemorrhoids (24.3%), angiodysplasia (25.0%), and nonspecific colitis (28.6%) were associated with intermediate admission rates, whereas diverticulosis (11.1%), normal colonoscopy (16.7%), and polyps (6.7%) had the lowest admission rates.

A comparable pattern was observed for the association between colonoscopic diagnosis and blood transfusion requirements (χ²[9] = 52.305; Monte Carlo exact p < 0.001). Malignant growth necessitated transfusion in 60.0% of affected patients (6/10), constituting by far the highest transfusion rate across all diagnostic categories. Solitary rectal ulcers and colitis were associated with transfusions in 20.0% (2/10) and 13.3% (2/15) of the patients, respectively. Only 1.4% of patients with hemorrhoids required transfusions (1/70), and no transfusions were administered to patients with angiodysplasia, diverticulosis, nonspecific colitis, polyps, normal colonoscopy, or stercoral ulcers. Both association analyses are summarized in Table 2.

6. Logistic Regression: Predictors of Hospital Admission

Hospital admission was modeled by binary logistic regression with three covariates: age, hemoglobin level, and colonoscopic diagnosis (entered as one coded variable). The model was statistically significant overall (likelihood ratio χ²[3] = 8.665, p = 0.034; McFadden pseudo-R² = 0.044). However, the Hosmer–Lemeshow test (χ²(8) = 104.47, p < 0.001) indicated poor calibration; thus, the estimates from this model need cautious interpretation.

Among the three covariates, only the hemoglobin concentration was independently and significantly associated with admission (odds ratio [OR] = 0.768; 95% confidence interval [CI]: 0.634–0.931; p = 0.007): every 1 g/dL increase in hemoglobin decreased the odds of admission by 23.2%. Neither age (OR = 1.002; 95% CI: 0.978–1.027; p = 0.866) nor the coded colonoscopic diagnosis variable (OR = 0.948; 95% CI: 0.813–1.105; p = 0.494) was independently associated with admission. Table 4 lists the odds ratios and confidence intervals for all the covariates.

Table 4. Binary Logistic Regression Analysis of Predictors of Hospital Admission.

Note. Outcome: hospital admission (reference category: not admitted). Model fit: likelihood ratio χ²(3) = 8.665, p = 0.034; Hosmer–Lemeshow χ²(8) = 104.47, p < 0.001; McFadden pseudo-R² = 0.044. OR = odds ratio; CI = confidence interval.

This study aimed to determine the clinical outcomes, colonoscopic findings, and endoscopic interventions in patients with LGIB in a resource-limited health care environment. The results revealed a heterogeneous spectrum of disease, with hemorrhoids being the most common diagnosis, followed by inflammatory conditions, polyps, vascular lesions, and malignancy. This pattern reflects the wide range of possible causes of LGIB and emphasizes the role of colonoscopy as an important diagnostic tool for the assessment and management of LGIB. Almost half of the patients in this study had hemorrhoids, which is consistent with previous studies reporting that anorectal disease was the predominant cause of LGIB, especially in younger and middle-aged individuals. Although very common, hemorrhoidal bleeding was not associated with high rates of hospital admission or transfusion, reflecting the relatively benign nature of the bleeding. In contrast, malignant lesions, although less common, were strongly associated with adverse clinical outcomes, including universal hospitalization and the highest transfusion rate. These findings are in line with the findings of previous studies showing that colorectal cancer patients who present with LGIB are likely to be at an advanced stage and to have substantial chronic occult blood loss and moderate-to-severe anemia [10,11]. The study also revealed that inflammatory and ulcerative conditions, such as colitis and solitary rectal ulcer syndrome, were associated with moderate-to-severe clinical disease, including increased hospitalization rates and transfusion requirements. These results corroborate the findings of previous studies showing that inflammatory bowel disease is a major source of LGIB-related morbidity because of persistent bleeding and mucosal ulceration [12,13,14]. In contrast, angiodysplasia and diverticulosis were associated with a relatively mild clinical course, with no transfusions needed, and occurred predominantly in older patients, reflecting age-related degenerative vascular and structural changes. One of the most important findings of this study was the difference in hemoglobin levels across diagnostic categories. The mean hemoglobin concentration was lowest in patients with colorectal malignancy, suggesting more severe and prolonged blood loss before presentation. These findings are consistent with previous literature indicating that bleeding associated with malignancy tends to be chronic and clinically insidious until late stages. Conversely, patients with a normal colonoscopy or nonspecific colitis had higher hemoglobin levels, suggesting less severe bleeding episodes or self-limiting disease [14,15,16]. Importantly, logistic regression analysis revealed hemoglobin concentration as the only independent predictor of hospital admission, making it a simple and practical marker of clinical severity in LGIB patients. The age differences observed across diagnostic groups in this study were consistent with those reported in previous epidemiological studies. Older patients more often present with diverticulosis, angiodysplasia, and malignancy, whereas younger patients present mostly with hemorrhoids and inflammatory conditions [17]. These results confirm the well-established age-related distribution of LGIB etiologies and underscore the importance of age in clinical risk stratification and diagnostic prioritization. In the current study, the pattern of endoscopic interventions reflects both the diagnostic and therapeutic uses of colonoscopy. Tissue biopsy was the most common procedure, highlighting the role of histopathology in inflammatory and neoplastic diseases. Polypectomy was successfully used to treat colonic polyps, enabling histological diagnosis and preventing potential malignant transformation. In selected cases, especially malignancies, endoscopic tattooing assisted surgical planning; although electrocautery hemostasis was rarely needed, it demonstrated the ability of colonoscopy to provide immediate therapeutic control of active bleeding. The rate of endoscopic hemostasis in this study (1.3%) was lower than the 7.4% endoscopic therapy rate reported in a population-based cohort from a high-resource setting [10], which may be attributable to delayed presentation, limited availability of advanced endoscopic equipment, and other resource constraints. The close relationship between endoscopic diagnosis and clinical outcomes, such as hospitalization and transfusion requirements, further confirms the prognostic value of endoscopic diagnosis. Patients with malignant or more severe inflammatory disease consistently had greater health care utilization, whereas patients with benign pathologies, such as hemorrhoids or polyps, were more likely to be managed as outpatients. These results support earlier research showing that endoscopic diagnosis is important for patient risk stratification and clinical decision-making [18,19]. Overall, this study revealed that the most common cause of LGIB in a resource-limited setting is benign anorectal disease, although a substantial proportion of patients have high-risk disease requiring intensive management. The findings also highlight the significance of hemoglobin as an important clinical marker of disease severity and hospitalization. Compared with reports from high-income settings, the results show some parallels in disease distribution as well as notable differences that may be attributable to variations in access to health care, diagnostic delay, and resources. This study has several limitations. First, its retrospective, single-center design may limit the generalizability of the findings and make it subject to selection and information bias. Second, several diagnostic subgroups were small, which reduced the statistical power of the subgroup comparisons. Third, important variables, such as comorbidities, antithrombotic medication use, the timing of colonoscopy, rebleeding, and mortality, were not reported. Finally, the logistic regression model showed poor calibration and included colonoscopic diagnosis as a single coded variable rather than as a separate diagnostic category; thus, the findings should be interpreted with caution. Prospective multicenter studies are needed to confirm these results.

LGIB in this resource-limited setting was predominantly caused by benign anorectal disease; however, a substantial proportion of patients had high-risk pathology, particularly for malignancy and colitis, and needed hospitalization and transfusion. Lower hemoglobin levels were independently associated with hospital admission and may serve as a simple marker of clinical severity. Colonoscopy remains essential for accurate diagnosis and targeted management, highlighting the need to strengthen endoscopic services in resource-constrained health care systems.