Self-Care Knowledge and Associated Factors Among Adult Patients With Colostomy in Northwest Amhara, Ethiopia: A Cross-Sectional Study

Abstract

Background

Colostomy, a surgically necessary procedure, requires rigorous self-care to prevent complications and improve patients’ quality of life. In Ethiopia, the rising demand for colostomies is met with a lack of localized data on patient self-care knowledge, leading to insufficient support and increased stoma-related complications and healthcare burdens.

Objectives

This study aimed to assess self-care knowledge and its associated factors among adult patients with colostomy in Northwest Amhara Comprehensive Specialized Hospitals, Ethiopia October-November 2024.

Methods

A hospital-based cross-sectional study involving 423 adult patients (aged ≥18) who were consecutively recruited from five hospitals (October 15– November 10, 2024) was conducted. Data were collected via structured interviews. Multivariable binary logistic regression was employed to identify factors associated with knowledge. Because participants were recruited from five hospitals, hospital site was considered as a potential clustering variable; its effect was examined in sensitivity analyses.

Results

Only 39.24% (95% CI: 34.57–43.92) of the participants (n=423, 68.3% male) demonstrated good colostomy self-care knowledge. Factors significantly associated with better knowledge included colostomy education (AOR = 8.72; 95% CI: 4.58–16.60), higher educational status (Elementary AOR = 3.48; Secondary AOR = 2.58; College and above AOR = 2.04), private employment (AOR = 3.45), and medium social support (AOR = 2.48). Government employment was associated with lower odds of good knowledge (AOR = 0.38). In a sensitivity analysis using stricter knowledge cut-off (≥6 of 9 items correct), only secondary education, colostomy education, and medium social support remained statistically significant, indicating that the findings for occupation and the remaining educational categories were not stable across alternative cut-offs.

Conclusion

Despite 75.4% of patients receiving preoperative teaching and 60.76% receiving colostomy-specific education, the prevalence of good self-care knowledge is suboptimal. Educational status, colostomy-specific education, and social support are significantly associated with self-care knowledge; however, only secondary education, colostomy education, and medium social support remain robust across alternative cut-offs, whereas the associations of occupation and other educational categories depended on the chosen threshold. These findings necessitate targeted, culturally sensitive educational programs to bridge knowledge gaps and improve patient outcomes in this resource-limited setting.

Introduction

Colostomy is a surgical procedure that creates an opening (stoma) in the large intestine and is essential for managing patients suffering from various illnesses, such as colorectal cancer, inflammatory bowel diseases, and even traumatic injuries (Nazarali & Gray, 2013; Recalla et al., 2013). Relocation of the colon to an external pouch can greatly ease some palliative symptoms and improve patients’ overall quality of life. However, colostomy requires extensive lifestyle alterations and rigorous self-care and may have both negative and positive impacts on patients (Al-beity, 2020; Sabea & Shaqueer, 2021).

Contemporary research, clinical practice, and even the most recent WCET standard edition focus on the importance of proficient self-care for stoma patients to avoid complications such as infections, stoma prolapse, or skin irritations that could result in increased morbidity and increased healthcare costs (Chabal et al., 2021). The range of stoma complications is immensely broad, with reports indicating stoma-related complications ranging from 20% to 80%, which highlights the critical importance of these issues (de Oliveira et al., 2018; Tiruneh et al., 2022).

Although extremely beneficial, colostomy can lead to serious physical, social, or mental challenges for patients. Issues such as fecal incontinence, flatus seepage, and changes in a person’s body image can lead to feelings of anxiety, helplessness, and social withdrawal (Kang & Choudhary, 2024; Sasaki et al., 2021). The procedure impacts personal interactions and normal routines, whether they are intended as temporary or permanent, with stigmatization and embarrassment being common problems faced by patients (Paszynska et al., 2023). While direct data on Ethiopian patients with colostomy self-care skills are lacking, available research shows that both patients and healthcare professionals often have insufficient knowledge (Atta & Mohammed, 2022; Belay et al., 2023; Sabea & Shaqueer, 2021).

Review of Literature

Knowledge Levels

Self-care knowledge among patients with colostomy is a multidimensional construct that encompasses understanding of stoma management, dietary modification, physical activities, psychosocial adaptation, and complication recognition. Prior research consistently demonstrates inadequate colostomy self-care knowledge across diverse settings. According to a study conducted in Nepal, knowledge regarding colostomy care, including normal function and complications, was found to be inadequate, with only 38.3% of participants demonstrating sufficient understanding (Pandey et al., 2015). Similarly, a study in Pakistan revealed that 50% of participants possessed poor knowledge regarding colostomy care (Hussain et al., 2020). A multi-center survey across the Philippines, Malawi, Nigeria, and India reported that 62.1% of participants had poor knowledge, with frequent difficulties in applying appliances correctly (Black, 2021). Furthermore, research in Vietnam indicated that 45.9% of participants had poor self-care knowledge (Ngo et al., 2023), while a study conducted in India reported that 76.7% of participants lacked knowledge about intestinal stoma self-care (Kadam & Shinde, 2014).

Self-Care Practices

Evidence on self-care practices among patients with colostomy further underscores substantial gaps. A study in Nepal showed that 7.4% of respondents had poor daily care practices, and the majority (72.3%) suffered from peristomal skin irritation (Pandey et al., 2015). In Southern Vietnam, 43.2% of participants demonstrated poor self-care practices for intestinal stomas (Ngo et al., 2023). Additionally, a study in Pakistan found that while 66.7% of participants washed their hands before procedures, 33.3% did not (Hussain et al., 2020).

Associated Factors and Quality of Life

Beyond knowledge and practice deficits, colostomy significantly affects patients’ psychosocial well-being and quality of life. A study in Poland found that one year following diagnosis, sexual activity decreased to 25% for women and 34% for men (Paszynska et al., 2023). A cross-sectional study in Brazil found that 30.89% of patients developed complications, with peristomal dermatitis being the most common (de Oliveira et al., 2018). Psychosocially, a Jordanian study revealed that while 45.8% of adults were independent in preparing equipment, approximately 40% were completely dependent on family for peristomal skin care (Ayik et al., 2021; Elshatarat et al., 2020). The aim of this study was (1) to assess the level of self-care knowledge among adult patients with colostomy and (2) to determine factors associated with good self-care knowledge.

Methods and Materials

Study Area and Period

The study took place in five Northwest Amhara Comprehensive Specialized Hospitals, namely Debre Markos, Tibebe Ghion, Debre Tabor, Felege Hiwot, and Gondar. These hospitals serve as major referral centers in the Amhara Region and provide surgical and follow-up services for patients with colostomy. Adult patients receiving care in the surgical wards and follow-up clinics were included. The study was conducted from October 15, 2024 to November 10, 2024.

Study Design

A prospective, hospital-based cross-sectional study design was employed to assess self-care knowledge and its associated factors among adult patients with colostomy. Data were collected through structured face-to-face interviews using a pretested questionnaire. It is important to acknowledge that consecutive sampling from a clinic-based population may not represent the entire colostomy patient population, particularly those with severe complications who may not attend follow-up appointments, or those who were not present during the data collection period, potentially leading to an overestimation of knowledge or an underestimation of complications.

Source Population

All adult patients with a colostomy who received care at the five Northwest Amhara Comprehensive Specialized Hospitals.

Study Population

All adult patients with a colostomy who visited surgical wards or follow-up clinics at the five Northwest Amhara Comprehensive Specialized Hospitals during the data collection period and met the inclusion criteria.

Inclusion and Exclusion Criteria

Inclusion Criteria

All adult patients aged 18 years and above with a colostomy who attended surgical wards or follow-up clinics and who were willing to participate were included.

Exclusion Criteria

Patients who were unwilling to participate and those with severe cognitive impairment.

Sample Size Determination

The sample size was determined using a single population proportion formula with the following assumptions: a 95% confidence level (Zα/2 = 1.96), a 5% margin of error, and a prevalence of good self-care knowledge of 50% (since no prior study in the Ethiopian context was available). Adding a 10% non-response rate, the final sample size was 423.

$$\mathrm{n}={\mathrm{Z}}^2\mathrm{P}\left(1-\mathrm{P}\right)/\hspace{0.17em}{\mathrm{d}}^2,\text{where},$$

$$=0.9604/0.0025=384$$

Where:

• N = minimum sample size

• Z = standard normal distribution (Z=1.96), CI of 95% = 0.05

• P = prevalence of knowledge; or practice; Hence, p=50%, % (0.5 will be used

• d = Absolute precision or tolerable margin of error= 5 % (0.05)

Therefore, based on the above assumption the minimum sample size is 384.

Then, by adding 10% of the non-response rate, the required and final adjusted sample size to be taken is 423 study subjects.

Sampling Technique

A consecutive sampling technique was employed to recruit participants. All eligible patients with colostomy who presented to the surgical outpatient departments and wards during the data collection period were approached and invited to participate until the required sample size was achieved. The five hospitals were included to ensure adequate representation and to reach the target sample size within the data collection period. See in detail (Figure 1).

Figure 1.

Sample technique for self-care knowledge, and associated factors among adult patients with colostomy at northwest Amhara comprehensive specialized hospitals, Ethiopia, 2024

Study Variables

Dependent Variable

• • Knowledge of colostomy self-care

Independent Variables

• • Sociodemographic characteristics (e.g., age, sex, residence, education level, occupation, marital status)

• • Health system factors (e.g., preoperative teaching, education on colostomy care, sources of information, previous hospitalization)

• • Psycho-social factors (e.g., social support, anxiety, depression, and body image disturbance).

Operational Definition

Self-care knowledge was assessed with nine items covering the nature of a normal colostomy, healthy stoma characteristics, signs of complications, and management of constipation and diarrhea. This knowledge assessment tool was adapted from previously published literature and modified for the local context (Collado-Boira et al., 2018). Each item was scored as correct (1 point) or incorrect (0 point). The nine items comprised: (1) nature of the colostomy (permanent vs. temporary); (2) normal color of a healthy stoma; (3) typical appearance of a healthy stoma; (4) signs of bleeding; (5) management of constipation; (6) management of diarrhea; (7) frequency of pouch change; (8) signs of complications; and (9) dietary considerations.

Good self-care knowledge was defined as a score of ≥50% (≥5 correct responses), while poor self-care knowledge was defined as a score <50%. This cut-off was chosen because there is no established universally accepted or clinically validated method to assess good versus poor self-care knowledge of colostomy, but such percentage-based classifications have frequently been used in knowledge assessment research (Abebe et al., 2019; Berhe et al., 2020; Beroza, 2024; Kebede et al., 2022). For items with multiple plausible responses (e.g., management of constipation), the correct-response coding rule used in the analysis is documented item-by-item in Appendix A; in brief, any single clinically recommended action was coded as correct, and “all of the above” was coded as correct only when it subsumed all clinically recommended options. The Results section further includes a sensitivity analysis using a higher cut-off of 6 out of 9 items correct (approximately 67% of the scale) to examine the robustness of the findings.

For the item assessing the nature of the colostomy (permanent vs. temporary), responses were verified against clinical records whenever available. If records were unavailable or incomplete, responses were judged against the documented surgical indication. When neither source was available, classification was based on self-report; this limitation is acknowledged in the Strengths and Limitations section.

The pouch-change item assessed knowledge of ostomy appliance replacement, not pouch emptying. During pretesting, data collectors were trained to clarify that the question referred to replacing the entire appliance to ensure consistent interpretation.

The knowledge assessment tool was adapted from (Collado-Boira et al., 2018) for the Ethiopian context through Amharic translation, cultural adaptation, and simplification of terminology. The complete instrument and scoring criteria are provided in Appendix A.

Psychometric Properties of Instruments

Social Support (SSRS)

Social Support was measured via the 12-item Social Support Rating Scale (Xiao, 1994). The SSRS measures three dimensions of social support: objective support, subjective support, and support utilization. The theoretical total score ranges from 0 to 66; however, in the present study the observed range was 0–20, consistent with the lower bound used in prior Chinese and Ethiopian applications of the scale. Scores were categorized as follows: Low support: 0–6; Medium support: 7–13; High support: 14–20 (Xiao, 1994). These tertile-style cut-offs correspond to the lower, middle, and upper thirds of the observed score distribution in the present sample and were chosen a priori to align with the original Chinese validation study’s recommended grouping. The discrepancy between the theoretical range (0–66) and the observed range (0–20) reflects the scoring structure of the SSRS as applied in the Ethiopian context. The original Chinese version of the SSRS uses a specific response format: items 1–4 and 6–8 are scored on a 4-point scale (1–4), items 5 and 9–10 are scored on a 3-point scale (1–3), and items 11–12 are scored on a dichotomous scale (0–1). The maximum total score under this scoring system is 44, and the observed range of 0–20 in the present study is consistent with other Ethiopian applications of the SSRS (Kassa et al., 2021; Xiao, 1994). The cut-offs of 0–6, 7–13, and 14–20 were established by (Xiao, 1994) based on the Chinese normative data and have been used in subsequent Ethiopian studies.

Anxiety and Depression Scale (HADS)

This instrument uses the 14-item Hospital Anxiety and Depression Scale (7 items for anxiety, 7 for depression). The HADS is a widely used screening tool that has been validated in multiple languages and cultural settings, including African populations (Snijkers et al., 2021). Each subscale yields a theoretical score of 0–21. Scores were categorized as follows: Presence of anxiety (HADS-A): Score ≥8; Presence of depression (HADS-D): Score ≥8 (Snijkers et al., 2021). In the Results, these categories are labelled “higher anxiety/depression” and “lower anxiety/depression,” where “higher” denotes a HADS score ≥ 8 (i.e., probable caseness) and “lower” denotes a HADS score < 8. The terms “higher” and “lower” were used to align with the analysis variable labels. “Higher” denotes probable clinical caseness (HADS ≥8), whereas “lower” indicates subthreshold symptoms (HADS <8); these terms distinguish probable cases from non-cases at the established screening cutoff and do not imply differences in symptom severity.

Body Image Disturbance

A 7-item survey adapted from the Body Image Disturbance Questionnaire (Cash et al., 2004), scored from 1 (never) to 5 (very often), was used. The scale was pre-tested and modified for cultural appropriateness in the Ethiopian context. The theoretical total score ranges from 7 to 35. Scores were categorized as follows: Lower disturbance: 7–21; Higher disturbance: 22–35 (Cash et al., 2004). These cut-offs represent the lower approximately 60% and upper approximately 40% of the theoretical range and were chosen to identify participants with above-median body-image concerns in a non-clinical colostomy population. These cut-offs are empirically derived rather than clinically validated; therefore, the resulting classification should be interpreted as a relative measure of body-image disturbance rather than an absolute clinical assessment. This limitation is acknowledged in the Strengths and Limitations section.

Data Collection Tool

Data were collected using a structured, interviewer-administered questionnaire through face-to-face interviews. The questionnaire consisted of four main sections: (1) sociodemographic characteristics (12 items); (2) health system factors such as preoperative teaching and colostomy education (8 items); (3) colostomy self-care knowledge (9 items); and (4) psychosocial factors assessed using the 12-item Social Support Rating Scale, 14-item Hospital Anxiety and Depression Scale (HADS), and a 7-item body image disturbance scale. The questionnaire was prepared in English and Amharic and pretested on 5% of the sample before use.

Reliability was assessed using Cronbach’s alpha computed from the full sample (n = 423) via SPSS v26; the 5% pretest (n = 21) provided only preliminary estimates. All five instruments demonstrated acceptable-to-good internal consistency. Colostomy knowledge (α = 0.78), SSRS (α = 0.82), HADS anxiety subscale (α = 0.85), HADS depression subscale (α = 0.87), and body image disturbance (α = 0.76). Corrected item total correlations ranged from 0.31 - 0.54 (knowledge), 0.34 - 0.61 (SSRS), 0.52 - 0.72 (HADS-A), 0.56–0.78 (HADS-D), and 0.33 - 0.58 (body image); all items exceeded the r ≥ 0.30 threshold, and no item removal increased any scale’s alpha by more than 0.02, so all items were retained.

Data Collection Procedure

Data were collected from October 15 to November 10, 2024, by five trained BSc nurses who were recruited from hospitals outside the study sites to avoid bias. Two MSc-level supervisors oversaw the process. The principal investigator provided daily coordination and on-site support to both data collectors and supervisors. Data collection was completed within 27 days because all five hospitals were sampled concurrently rather than sequentially, with one trained data collector stationed at each study site throughout the data-collection period, enabling recruitment of the planned sample of 423 participants.

Eligible patients attending the surgical wards or follow-up clinics were approached consecutively. After explaining the study purpose, benefits, and risks in a private setting, verbal informed consent was obtained from those willing to participate. All approached patients who met the inclusion criteria agreed to participate, resulting in a 100% response rate. Face-to-face interviews were then conducted using the structured questionnaire in either Amharic or English, according to participant preference. Each interview lasted approximately 25–35 minutes and was completed in a single session. Completed questionnaires were checked for completeness on-site, and any missing responses were addressed before concluding the interview.

A participant flow summary is presented in Figure 1. Of 445 patients with colostomy screened across five hospitals, 432 met the eligibility criteria. Nine patients were excluded due to unwillingness or inability to provide informed consent, and five patients were excluded because they were unable to complete the interview due to acute illness. All 423 eligible patients approached agreed to participate, resulting in a 100% response rate. Participants were distributed across Debre Markos (n = 74), Tibebe Ghion (n = 111), Debre Tabor (n = 67), Felege Hiwot (n = 89), and Gondar (n = 82).

Ethical Considerations

Ethical clearance was granted by the Institutional Review Board of Debre-Markos University Approval number [RCSTTD/282/01/17], on September 13, 2024. Written permission was also received from each of the five selected hospitals. Informed consent was obtained from all participants before data collection after explaining the purpose of the study, study procedures, benefits, risks, and any queries they may have had. Participation was voluntary, and participants had the right to withdraw at any time without penalty. Privacy and confidentiality were assured by encoding responses with unique numbering codes instead of using names, and all information was securely stored and protected from unauthorized access.

Data Quality Assurance

To ensure data quality, a pretest was conducted on 5% of the sample size (21 participants) at Finote Selam General Hospital before the actual data collection. The questionnaire was originally prepared in English, translated into Amharic by a language expert, and then back translated into English to ensure consistency and accuracy.

Five data collectors and two supervisors received half-day training on the study objectives, ethical considerations, confidentiality of information, participants’ rights, and proper interviewing techniques. During data collection, the principal investigator and supervisors provided daily on-site supervision. Completed questionnaires were checked for completeness and accuracy immediately after each interview, and any inconsistencies or missing data were addressed on the spot. All collected data were securely stored and kept confidential.

Data Analysis and Interpretation

The collected data were checked for completeness and consistency by the principal investigator on a daily basis. Data were then coded, entered, and cleaned using EpiData version 4.4.0 software and exported to SPSS version 26 for analysis. Descriptive statistics (frequencies, percentages, means, and standard deviations) were used to summarize sociodemographic characteristics, health system factors, psychosocial factors, and levels of self-care knowledge. Results were presented using tables, figures, and text.

Bivariate logistic regression was first performed to examine crude associations. Independent variables with p < .25 in the bivariate analysis were considered as candidates for multivariable analysis. This more inclusive threshold, rather than the conventional p < .05 or a stricter p < .01, was chosen to reduce the risk of excluding potentially important confounders at an early stage, thereby supporting a more comprehensively adjusted multivariable model. Although a p < .25 cutoff may include some variables that are not individually significant, this approach is well-established in the epidemiological literature (Hosmer et al., 2013) and minimizes the risk of omitted-variable bias while the final model still relies on the adjusted odds ratios and their significance levels to identify true associations.

These variables were entered into multivariable binary logistic regression to control for confounding. Adjusted odds ratios (AOR) with 95% confidence intervals (95% CI) were computed to identify factors independently associated with good colostomy self-care knowledge. Multicollinearity was checked using Variance Inflation Factor (VIF), with all VIF values ranging from 1.16 to 1.72 (well below the threshold of 10), indicating no significant multicollinearity. The Hosmer–Lemeshow goodness-of-fit test was non-significant, χ² (8) = 3.80, p = .875, indicating good model fit.

Because participants were recruited from five hospitals, the analysis considered possible hospital-level clustering. Hospital site was included as a fixed effect in an alternative model specification; the direction, magnitude, and significance of the associated factors were unchanged, and the intraclass correlation coefficient for the knowledge outcome across hospitals was negligible (ICC < 0.01), indicating that single-level logistic regression was appropriate. The results presented in Table 5 therefore report the single-level multivariable model without hospital fixed effects. A sensitivity analysis was also performed using a stricter knowledge cut-off (≥6 of 9 items correct, approximately 67%) to examine the robustness of the identified associations.

Table 5.

Multivariable Logistic Regression Analysis of Associated Factors Among Patients Who Underwent Colostomy at Northwest Amhara Comprehensive Specialized Hospitals, Ethiopia, 2024

Variable	Category	Knowledge
		Good (%)	Poor (%)	COR (95% CI)	AOR (95% CI)	Value
Age	18-29	13 (3.07)	27 (6.40)	2.41 (0.85-6.87)	1.62 (0.48-5.47)	.437
	30-49	72 (17.02)	105 (24.80)	3.43 (1.44-8.16)	1.34 (0.45-3.98)	.599
	50-69	74 (17.50)	90 (21.30)	4.11 (1.72-9.82)	1.38 (0.47-4.05)	.558
	≥ 70	7 (1.65)	35 (8.30)	1	1	1
Educational status	Unable to read & write	46 (10.90)	115 (27.20)	1	1	1
	Read & write	22 (5.20)	50 (11.80)	1.10 (0.61-2.02)	1.12 (0.55-2.28)	.755
	Elementary	21 (5.00)	16 (3.80)	3.28 (1.58-6.83)	3.48 (1.42-8.52)	.006
	Secondary	32 (7.62)	24 (5.70)	3.33 (1.80–6.19)	2.58 (1.24-5.38)	.011
	College and above	45 (10.60)	52 (12.30)	2.16 (1.28–3.66)	2.04 (1.08-3.86)	.028
Occupation	Government employment	24 (5.70)	55 (13.00)	0.85 (0.47–1.56)	0.38 (0.17-0.86)	.019
	Private employment	30 (7.10)	14 (3.30)	4.19 (2.06–8.53)	3.45 (1.48-8.04)	.004
	Merchant	22 (5.20)	30 (7.10)	1.43 (0.77–2.65)	1.28 (0.58-2.82)	.541
	Student	4 (0.90)	3 (0.70)	2.61 (0.57–11.95)	3.15 (0.62-16.04)	.167
	Housewife	17 (4.00)	20 (4.70)	1.66 (0.83–3.31)	1.54 (0.62-3.82)	.352
	Farmer	69 (16.30)	135 (31.90)	1	1	1
Previous Hospitalization	Yes	104 (24.60)	145 (34.30)	1.29 (0.87-1.93)	1.18 (0.73-1.91)	.500
	No	62 (14.70)	112 (26.50)	1	1	1
Education on colostomy care	Yes	148 (34.99)	109 (25.80)	11.16 (6.44–19.32)	8.72 (4.58-16.60)	<.001
	No	18 (4.26)	148 (34.99)	1	1	1
Social support	Lower social support	35 (8.30)	90 (21.30)	1	1	1
	Medium social support	64 (15.10)	50 (11.80)	3.29 (1.92-5.64)	2.48 (1.28-4.80)	.007
	High social support	67 (15.80)	117 (27.70)	1.47 (0.90-2.41)	1.12 (0.62-2.02)	.707

Results

Sociodemographic Factors of Study Participants

A total of 423 adults who underwent colostomy participated in the study, resulting in a 100% response rate achieved through consecutive enrollment. The participants had a mean age of 49 years (SD = 14.50), with ages ranging from 18 to 70 years. Most participants were in the 30–49-year age group, male, and residents of rural areas. A substantial proportion were unable to read and write, and farming was the predominant occupation. Detailed sociodemographic characteristics of the study participants are presented in Table 1.

Table 1.

Distribution of Sociodemographic Characteristics of Adult Patients Who Underwent Colostomy at Northwest Amhara Comprehensive Specialized Hospitals, Ethiopia, 2024 (n = 423)

Category	Subcategory	Frequency	Percent
Age	18-29	40	9.50
	30-49	177	41.80
	50-69	164	38.80
	≥70	42	9.90
Gender	Male	289	68.30
	Female	134	31.70
Residence	Urban	183	43.30
	Rural	240	56.70
Education	Unable to read and write	161	38.06
	Read and write	72	17.00
	Elementary	37	8.70
	Secondary	56	13.24
	College and higher	97	22.90
Marital Status	Single	125	29.60
	Married	179	42.30
	Divorced	46	10.90
	Widowed	42	9.90
	Separated	31	7.30
Religion	Orthodox	306	72.30
	Muslim	87	20.60
	Protestant	26	6.10
	Other	4	0.90
Occupation	Governmental employer	79	18.70
	Private employer	44	10.40
	Merchant	52	12.30
	Farmer	204	48.20
	Housewife	37	8.70
	Student	7	1.70

Health System and Psychosocial Factors

Health system factors influencing colostomy care were examined. Preoperative teaching was received by 319 participants (75.40%), and 257 (60.76%) received education on colostomy care. The primary information sources included healthcare professionals (224 participants, 52.96%) and the internet (33 participants, 7.80%). The information-source responses exceeded the number of participants who received colostomy education because participants could select multiple information sources, whereas colostomy education was assessed as a binary (yes/no) variable. Previous hospitalization was reported by 249 participants (58.87%). See Table 2 for details.

Table 2.

Frequency Distribution of Health System Characteristics of Adult Patients Who Underwent Colostomy at Northwest Amhara Comprehensive Specialized Hospitals, Ethiopia, 2024 (n = 423)

Question	Category	Frequency	Percent
Preoperative Teaching	No	104	24.60
	Yes	319	75.40
Education on Colostomy Care	No	166	39.24
	Yes	257	60.76
Sources of Colostomy Care Information	Televisions	13	3.07
	Journals	12	2.84
	Internet	33	7.80
	Healthcare Professional	224	52.96
	Others	15	3.55
Previous Hospitalization	No	174	41.10
	Yes	249	58.90
Number of Hospitalizations	First Time	48	11.30
	Second Time	39	9.20
	Three Times	38	9.00
	Four Times or More	124	29.30

Psychosocial factors were assessed using validated measurement scales (Table 3). Overall, 279 participants (66.0%) exhibited higher levels of anxiety, while 260 (61.5%) reported higher levels of depression. Regarding social support, 125 participants (29.6%) had lower social support, whereas 184 (43.5%) reported higher social support. Body-image disturbance was higher in 243 patients (57.45%).

Table 3.

Distribution of the Psychosocial Status of Patients Who Underwent Colostomy at Northwest Amhara Comprehensive Specialized Hospitals, Ethiopia, 2024 (n = 423)

Variable	Category	Frequency	Percent
Anxiety	Lower Anxiety	144	34.04
	Higher Anxiety	279	65.96
Depression	Lower Depression	163	38.50
	Higher Depression	260	61.50
Social support	Lower social support	125	29.60
	Medium social support	114	27.00
	Higher social support	184	43.50
Body image disturbance	Lower body image disturbance	180	42.55
​	Higher body image disturbance	243	57.45

Prevalence of Self Care Knowledge

The study assessed participants’ knowledge of colostomy self-care. Among the 423 participants, 166 (39.24%; 95% CI: 34.57–43.92) demonstrated good self-care knowledge, whereas 257 (60.76%) had poor knowledge. Figure 2 illustrates the distribution of knowledge levels among participants.

Figure 2.

Distribution of colostomy self-care knowledge levels (good knowledge vs. poor knowledge) among participants at Northwest Amhara comprehensive specialized hospitals, Ethiopia, 2024 (n = 423)

Item-level knowledge responses are presented in Table 4. Most participants (377, 89.1%) correctly identified red as the normal color of a healthy stoma; however, only 161 (38.1%) correctly recognized signs of bleeding, indicating a substantial gap in knowledge of stoma-related complications. Overall, 248 participants (58.6%) were aware of the nature of their colostomy. Regarding pouch management, 239 participants (56.5%) reported changing their colostomy pouch twice daily.

Table 4.

Participants’ Knowledge Related to Colostomy and Self-Care at Northwest Amhara Comprehensive Specialized Hospitals, Ethiopia, 2024 (n = 423)

Question	Response	Frequency	Percent
Do you know the nature of your colostomy?	False	175	41.40
	True	248	58.60
If it is true, which one is?	Permanent	101	23.90
	Temporary	147	34.80
Could you identify signs of bleeding from stoma?	False	262	61.90
	True	161	38.10
What is the normal color of a healthy colostomy?	Red	377	89.10
	Pale	30	7.10
	Black	8	1.90
	Don’t know	8	1.90
What does a typical colostomy look like?	Shiny and moist	119	28.10
	Has nodules	191	45.20
	Has cuts and torn	113	26.70
How do you manage constipation?	Increase fluid intake	85	20.10
	Use of stool softener	151	35.70
	Take medical help	178	42.10
	All of the above	9	2.10
How do you manage diarrhea?	Avoid food items that cause diarrhea	71	16.80
	Use over-the-counter drugs	87	20.60
	Seeking medical help	231	54.60
	All of the above	34	8.00
How often should the colostomy appliance be changed?	Every third day	10	2.40
	Weekly	3	0.70
	Every other day	6	1.40
	Daily	165	39.00
	Twice a day	239	56.50
Signs of complication	Recognized ≥ 3 of 5 signs	148	34.99
	Recognized < 3 of 5 signs	275	65.01
Dietary considerations+	Identified ≥ 3 of 5 items	179	42.32
	Identified < 3 of 5 items	244	57.68

Sensitivity Analysis of Knowledge

To explore the reliability of the results relative to the binary knowledge classification, we performed sensitivity analysis with a new cut-off of 6 out of 9 items correct (67%) to indicate good knowledge. This cut-off was calculated by multiplying 67% by the number of items (9). The analysis was based on 423 respondents. 89 participants (21.04%; 95% CI: 17.27-25.32) were considered to have good knowledge with the more conservative cut-off of 6 items correct compared with 166 (39.24%) with the mean-based cut-off used in the primary analysis. We reconducted the logistic regression analysis using this new cut-off of 6 items correct and found that 3 variables were significantly associated with good knowledge: preoperative colostomy education (AOR = 6.92; 95% CI: 3.41-14.04; p < .001), secondary education (AOR = 4.87; 95% CI: 1.98-11.96; p < .001), and medium social support (AOR = 2.08; 95% CI: 1.06-4.09; p = .034). With the stricter cut-off, fewer participants were classified as having good knowledge (89 [21.04%] vs. 166 [39.24%]).

Importantly, the set of associated factors was not identical under the two cut-offs. Of the seven categories that were significant in the primary model (elementary, secondary, and college education; government and private employment; colostomy education; and medium social support), only three secondary education, colostomy education, and medium social support remained statistically significant under the stricter cut-off. Occupation (both government and private employment) and the elementary and college educational categories lost statistical significance. These results indicate that the associations of occupation and the broader educational gradient with good knowledge are sensitive to the chosen cut-off, whereas the associations of secondary education, colostomy education, and medium social support are more robust. The findings underscore the importance of interpreting the knowledge prevalence in light of the chosen cut-off and support the recommendation that future studies employ more nuanced, multi-tiered classifications.

Factors Associated With Self-Care Knowledge

Factors associated with good self-care knowledge were analyzed via binary and multivariable logistic regression. In the multivariable model, educational status, occupation, prior colostomy education, and social support were significantly associated with knowledge (p ≤ .05).

Compared with participants unable to read or write, those with an elementary education had an adjusted odds ratio (AOR) of 3.48 (95% CI: 1.42–8.52, p = .006), those with a secondary education had an AOR of 2.58 (95% CI: 1.24–5.38, p = .011), and those with a college education or higher had an AOR of 2.04 (95% CI: 1.08–3.86, p = .028). Compared with farmers, government employees had an AOR of 0.38 (95% CI: 0.17–0.86, p = .019), while private employees had an AOR of 3.45 (95% CI: 1.48–8.04, p = .004). The participants who received colostomy education had an AOR of 8.72 (95% CI: 4.58–16.60, p < .001). Compared with lower support, medium social support was associated with an AOR of 2.48 (95% CI: 1.28–4.80, p = .007). The full results are shown in Table 5.

Discussion

This study assessed colostomy self-care knowledge and its associated factors among adults at Northwest Amhara Comprehensive Specialized Hospitals, Ethiopia. Despite relatively high rates of colostomy-specific education (60.76%) and preoperative teaching (75.42%), only 39.24% of the participants reported good knowledge. These findings highlight persistent gaps in patient understanding and the need for more effective, patient-centered educational strategies. As the first study of its kind in the Amhara region, this study contributes to the limited body of evidence on ostomy self-care in low-resource settings.

The demographic profile revealed a predominance of male participants (68.32%) with a mean age of 49 years, aligning with regional trends such as the 4:1 male-to-female ratio in Uganda (Ssewanyana et al., 2021) and similar male predominance in Kenya (Mungai et al., 2021). This pattern likely reflects the higher incidence of colostomy due to trauma, infectious conditions, and emergency surgeries, which are more common among younger males in sub-Saharan Africa (Abotchie et al., 2012; Salem et al., 2024; White et al., 2018). In contrast, high-income countries report older patient cohorts and malignancy as the leading indication for colostomy, with mean ages ranging from 60.60 to 65.50 years (Bosshardt, 2003; Romão et al., 2020; Skibsted et al., 2022). These demographic differences stem from varying healthcare infrastructures and disease burdens. While high-income countries primarily treat colorectal cancer in older adults, low-resource settings like Ethiopia report more trauma and emergency-related colostomies in younger males lacking preventive care.

Our findings on knowledge levels are lower than those from Vietnam (54.10%) and India (50.40%) (Kang & Choudhary, 2024; Ngo et al., 2023) but considerably higher than the 13.30% reported in Pakistan (Din et al., 2022). These international differences are not random; they reflect profound contextual variations in healthcare delivery, patient literacy, and post-operative support systems. In the Vietnamese hospital-based study, participants benefited from a more structured follow-up environment in southern Vietnam, where repeated encounters with stoma-care nurses and visual demonstration materials are more routine. Similarly, the Indian study was conducted among ostomates who often receive repeated counseling through dedicated cancer-care pathways and community support groups, which reinforce knowledge long after discharge.

In contrast, the extremely low knowledge level documented in Pakistan (13.30 %) occurred in a tertiary-care setting with comparable resource limitations to Ethiopia but possibly even weaker preoperative preparation and almost no routine follow-up. Within the Ethiopian context, the 39.24 % good-knowledge rate is remarkably consistent with the 38.3 % reported among Nepalese ostomates and the 37.9 % pooled good-knowledge rate from a multi-country survey in the Philippines, Malawi, Nigeria, and India (Lapitan et al., 2024; Pandey et al., 2015). These parallel findings across low- and middle-income countries (LMICs) underscore a systemic challenge: even when health professionals provide education (as was the case for 60.76% of our participants), the content is frequently delivered in a single session, without hands-on practice, pictorial aids, or follow-up reinforcement. In Northwest Amhara, 56.7 % of participants resided in rural areas, 38.3 % were unable to read or write, and 48.2 % were farmers demographic realities that make abstract verbal instructions difficult to internalize. Moreover, the predominance of trauma-related and emergency colostomies (reflected in the younger mean age of 49 years and 68.3 % male participants) means many patients receive their stoma unexpectedly, with limited psychological preparation compared with elective cancer patients in high-resource settings.

The strong positive association between educational attainment and knowledge further illuminates these mechanisms. Compared with participants unable to read or write, those with elementary, secondary, and college-level education had higher odds of good knowledge. This gradient mirrors findings from Taiwan, mainland China, and southern Egypt, where higher literacy directly translates into better comprehension of complication signs, dietary management, and pouch-changing schedules. In our setting, low literacy does not merely limit reading; it restricts the ability to ask clarifying questions, remember verbal instructions, or teach family members critical skills in a culture where family involvement in care is normative. However, this educational gradient was only partly robust across alternative knowledge cut-offs: in the sensitivity analysis using the stricter threshold (≥6 of 9 items correct), only secondary education remained statistically significant (AOR = 4.87; 95% CI: 1.98–11.96), whereas elementary and college-level associations did not retain significance, indicating that the broader educational gradient is partly threshold-dependent (Abdelmohsen, 2020; Cheng et al., 2013; Lo et al., 2011).

The participants who received colostomy-specific education were more than eight times more likely to have good knowledge, supporting similar conclusions drawn in Indonesia and Egypt, and this association remained robust under the stricter knowledge cut-off (AOR = 6.92; 95% CI: 3.41–14.04), supporting the stability of this finding across alternative thresholds (Leli Herawati, 2019; Moussa, 2019). These findings underscore the vital role of culturally sensitive, structured educational programs. In addition to education, occupational status was significantly associated with knowledge levels. Private-sector employees demonstrated better self-care than farmers (AOR = 3.45; 95% CI: 1.48-8.04; p = .004). The observed association is likely multifactorial, but the present study did not directly measure the underlying mechanisms; accordingly, the interpretations offered below should be regarded as hypotheses for future investigation rather than explanations established by the data. As working hypotheses, we propose that private employees may benefit from urban residence with greater proximity to healthcare facilities, more regular access to medical services, and potentially more flexible schedules that permit attendance at follow-up appointments, whereas farmers’ agricultural obligations and geographic isolation from urban hospitals may create barriers to continuity of care. It is also plausible but unmeasured in the present study that the indication for colostomy (e.g., elective versus emergency surgery; trauma-related versus disease-related) differs across occupational groups, which would in turn affect the opportunity for preoperative counseling. These hypotheses are consistent with studies from Egypt and Turkey (Sabah et al., 2017; Çulha et al., 2016), but require confirmation in future studies that directly measure indication for colostomy, employment-related healthcare access, and scheduling flexibility.

Government employees were 62% less likely to have good knowledge about colostomy self-care compared with farmers (AOR = 0.38; 95% CI: 0.17-0.86; p = .019). This counter-intuitive finding remained significant after adjusting for educational status, indicating that educational attainment alone does not explain the association. A possible explanation is that government employees in this setting have better structural access to formal healthcare and may therefore rely on clinicians for ongoing management, whereas farmers who have poorer access may be driven to learn about their condition through trial-and-error self-management; this interpretation has been observed in other Ethiopian chronic-disease studies (Kassa et al., 2021). However, this interpretation is offered as a working hypothesis rather than a finding established by the present data and may also reflect residual confounding by indication for colostomy, regional variation in the structuring of colostomy-education services (Tiruneh et al., 2022), or the small sample size in certain occupational categories, which limits the precision of the estimate. Furthermore, the inverse association of government employment was not robust across alternative knowledge cut-offs and lost statistical significance under the stricter threshold, indicating that this counter-intuitive finding should be interpreted with particular caution.

Social support was significantly associated with self-care knowledge, but the association was non-linear: only medium social support was significantly associated with good knowledge (AOR = 2.48; 95% CI: 1.28-4.80; p = .007) compared with lower support, whereas higher social support was not significantly associated with good knowledge (AOR = 1.12; 95% CI: 0.62-2.02; p = .707). This non-linear pattern suggests that a moderate level of social support may be most conducive to acquiring self-care knowledge, while very high support may sometimes reflect a patient who is more dependent on caregivers and therefore less personally engaged in learning stoma-care skills. Alternatively, the non-linear pattern may reflect unmeasured confounders (e.g., disease severity, presence of a primary caregiver, household structure) that warrant investigation in future studies. Notably, the association of medium social support with good knowledge remained robust under the stricter knowledge cut-off (AOR = 2.08; 95% CI: 1.06–4.09; p = .034), supporting the stability of this finding, whereas the non-significant association of high social support should be interpreted with caution given the sensitivity of the results to the chosen knowledge threshold. The findings therefore support targeted involvement of family members in education for patients with low support, while cautioning against assuming that more support is uniformly better.

Strengths and Limitations

The study conducted at five Northwest Amhara Comprehensive Specialized Hospitals provided a crucial assessment of self-care knowledge among patients with colostomy, serving as the first of its kind in the region to address evidence gaps in this resource-limited setting. This research highlights key associated factors, such as education and social support, offering a foundation for designing targeted patient management strategies. However, several limitations should be noted. The cross-sectional design does not allow causal inference or assessment of temporal relationships; accordingly, the reported associations should not be interpreted as causal effects.

Consecutive sampling from a clinic-based population may limit representativeness, as regular follow-up patients may differ from others; specifically, patients who were not present during the data collection period including those with severe complications, those who missed follow-up appointments, or those who had died were not included, potentially leading to an overestimation of knowledge or an underestimation of complications. Self-reported data are subject to response, recall, and social desirability bias. Although the knowledge tool showed acceptable reliability, full psychometric validation in the Ethiopian population was not performed.

Additionally, the binary classification of knowledge using a ≥ mean cut-off on a 9-item questionnaire, while consistent with prior literature, may not adequately capture clinically meaningful differences in knowledge. A patient scoring at the threshold (5 out of 9) may still lack critical knowledge about serious complications such as parastomal hernia or severe skin breakdown. Future studies should consider employing a more nuanced categorization (e.g., poor, moderate, good). The study was conducted in referral hospitals serving mainly rural populations, which limits generalizability to urban areas and other regions. Direct observation of self-care practices was not conducted, reducing objective verification of reported knowledge. Further limitations should be considered. The assessment of colostomy type (permanent vs. temporary) relied on clinical records when available and self-report when records were unavailable, which may have introduced misclassification bias. The body-image disturbance cut-offs were empirically derived and should therefore be interpreted as relative rather than clinically validated measures. Although the 100% response rate reflected the clinical setting and consecutive sampling approach, potential concerns regarding voluntary participation and perceived pressure from healthcare providers cannot be completely excluded. Additionally, the study did not collect data on the indication for colostomy (e.g., elective vs. emergency, trauma-related vs. disease-related), which may limit interpretation of occupational and demographic associations due to possible confounding effects.

Clinical Implication

The clinical implications highlight the urgent need for enhanced culturally appropriate education programs. With only 39.24% demonstrating good knowledge despite high education rates, clinicians should focus on improving content delivery. Structured interventions could bridge knowledge gaps (Chauhan & Jindal, 2017; Hegazy et al., 2014). Targeting younger, male-dominated cohorts affected by trauma with accessible, practical training could also address regional needs.

Nursing Implication

The findings highlight the need for a structured, continuous nursing care model rather than one-time patient education.

Standardized Preoperative Education

Despite widespread preoperative teaching, knowledge levels remained low. Colostomy-specific education was significantly associated with better knowledge (AOR = 8.72; 95% CI: 4.58–16.60) and remained robust under the stricter knowledge cut-off (AOR = 6.92; 95% CI: 3.41–14.04), emphasizing the need for mandatory, practical training on diet, pouch management, and complication recognition.

Literacy-Sensitive Teaching Approaches

Given high rates of illiteracy and the strong association between education level and knowledge, written materials alone are insufficient. However, nurses should be aware that the educational gradient was only partly robust across alternative knowledge cut-offs: only secondary education remained significantly associated with good knowledge under the stricter threshold, whereas elementary and college-level associations were sensitive to the chosen cut-off. Nurses should use pictorial guides, hands-on demonstrations, anatomical models, and the teach-back method to improve understanding of stoma care and complications.

Risk-Based Discharge Planning

Patients with greater vulnerability, particularly those from rural areas or agricultural occupations, should receive enhanced discharge support and closer follow-up.

Psychosocial and Continuity Support

Only medium social support was significantly associated with better self-care knowledge (AOR = 2.48), while higher social support was not (AOR = 1.12; p = .707), nurses should assess the patient’s existing support network and actively involve family members in education, particularly for patients with low support. Notably, the association of medium social support with better knowledge remained robust under the stricter knowledge cut-off (AOR = 2.08; 95% CI: 1.06–4.09; p = .034), whereas the non-significant association of high social support should be interpreted with caution given the sensitivity of the results to the chosen knowledge threshold. The non-linear pattern observed here suggests that interventions should aim to ensure at least moderate engagement of caregivers rather than assuming that more support is uniformly better.

Conclusion and Recommendations

This study underscores three key points. The sociodemographic profile of patients with colostomy in the Amhara region is characterized by older age, male predominance, and low educational attainment; self-care knowledge is suboptimal, with only 39.24% of patients demonstrating good understanding; and several factors are significantly associated with self-care knowledge, including educational status, colostomy-specific education, occupation, and social support. Importantly, of these, only secondary education, colostomy-specific education, and medium social support remained robust across alternative knowledge cut-offs, whereas the associations of elementary and college education and of occupation depended on the chosen threshold.

These insights highlight the need for targeted educational interventions and strengthened support systems to address knowledge gaps and improve patient outcomes. Future research should prioritize longitudinal studies to further elucidate the temporal and potentially causal nature of the observed associations, as the cross-sectional design of the present study precludes causal inference. In interpreting the present findings, readers should note that the associations of secondary education, colostomy-specific education, and medium social support were robust across alternative knowledge cut-offs, whereas the associations of elementary and college education and of occupation were sensitive to the chosen threshold and therefore warrant cautious interpretation. Furthermore, researchers should replace binary knowledge cut-offs with nuanced, multi-tiered classifications (e.g., poor, moderate, good) or sensitivity analyses.

ORCID iDs

Berhanu Ayana Tafere https://orcid.org/0009-0005-8852-3491

Benalfew Mengistie Lake https://orcid.org/0000-0003-1020-5678

Ethical Considerations

Ethical clearance letter was granted by the Institutional Review Board of Debre-Markos University Approval number [RCSTTD/282/01/17] on [September 13, 2024]. Written permission letter was also received from each of the five selected hospitals respectively. Verbal consent was taken from all participants before the data collection exercise after explaining to them the purpose of the study, study procedures, benefits, risk and any query they may have. Participation in the study was voluntary, and they had the right to withdraw at any time without penalty. Privacy and confidentiality was assured by encoding their responses with unique numbering codes instead of using names and all information was securely kept hidden from unauthorized people.

Data Availability Statement

The data used and/or analyzed during the current study are available from the corresponding author on reasonable request.*