Access to safe sanitation services and factors associated with personal protective equipment utilisation among sanitation workers in Ethiopia: a mixed-methods study

Abstract

Objective

Sanitation workers experience elevated occupational health risks due to hazardous exposures and social marginalisation. This study assessed access to safe sanitation services and examined Health Belief Model-based factors associated with personal protective equipment (PPE) use among sanitation workers in Ethiopia.

Design

A concurrent mixed-methods cross-sectional study was conducted from November 2021 to June 2022. Quantitative data were collected through face-to-face interviews using a structured interviewer-administered questionnaire via multistage simple random sampling. Concurrently, qualitative data were collected through key informant interviews and focus group discussions guided by semistructured interview guides. The quantitative data were analysed using Stata V.14, and associations between variables were examined using multivariable logistic regression. The results were presented as adjusted ORs with 95% CIs, and statistical significance was set at p<0.05. Qualitative data were analysed thematically using OpenCode V.4.03.

Setting

The study was conducted in Addis Ababa, Ethiopia.

Participants

The quantitative survey included 821 sanitation workers, while the qualitative component comprised 12 key informants and 65 focus group participants.

Outcome measures

The primary outcomes were access to safe sanitation services and PPE use.

Results

Overall, 38.2% (95% CI 34.9% to 41.6%) of sanitation workers had adequate access to safe sanitation services and 54.3% (95% CI 50.8% to 57.8%) reported PPE use. PPE use was positively associated with larger household size, workplace PPE availability, supervision, higher cues to action and greater self-efficacy, whereas perceived barriers were associated with lower odds of PPE use. Qualitative findings complemented these results by showing that inconsistent PPE provision, inadequate occupational health training, weak enforcement of safety regulations, precarious employment conditions and persistent social stigma contributed to poor occupational safety practices.

Conclusion

Sanitation workers in Addis Ababa have limited access to safe sanitation services and suboptimal PPE use. Improving organisational support, consistent PPE provision, effective supervision and occupational health training may enhance PPE use and strengthen the health, safety and resilience of this essential workforce.

STRENGTHS AND LIMITATIONS OF THIS STUDY.

• This study employs a concurrent mixed-methods design, enabling triangulation of quantitative and qualitative findings.

• This study uses multistage random sampling and a large sample to strengthen the robustness of the quantitative findings.

• This study applies the Health Belief Model as a theory-informed framework to examine factors associated with personal protective equipment utilisation.

• Self-reported measures and dichotomisation of Health Belief Model constructs may have introduced reporting bias and reduced measurement precision.

• The cross-sectional quantitative design precludes causal inference, and the observed relationships should be interpreted as associations.

Introduction

Access to basic sanitation is a fundamental component of global development agendas, including the former Millennium Development Goals and the current Sustainable Development Goals (SDGs), particularly Goal 6 (Clean Water and Sanitation) and Goal 8 (Decent Work and Economic Growth).^{1 2} Sanitation workers, tasked with managing human waste and maintaining public cleanliness, are indispensable to realising these goals. However, in many low- and middle-income countries (LMICs), including Ethiopia, these workers operate in hazardous environments with limited legal protections, recognition or support.^3–8 Their roles often remain invisible to policy and regulatory frameworks, leaving them disproportionately vulnerable to exploitation and harm.^{7 9} Despite their crucial contributions to public health, sanitation workers remain among the most marginalised and overlooked sectors, both in terms of research and the provision of essential occupational protections.¹⁰ Global evidence shows that failure to integrate sanitation worker safety into health system frameworks can impede progress on universal health coverage and sustainable urban living.¹¹

Sanitation workers are regularly exposed to biological, chemical and physical hazards that increase their risk of occupational injury, illness and even death.¹² Many work informally without access to adequate training, hygienic facilities or personal protective equipment (PPE), and they frequently lack preventive healthcare, medical treatment and social protections.^13–15 They also endure significant social stigma, discrimination and financial insecurity, further compounding their marginalisation and vulnerability.^{4 7 16} The failure to adequately protect sanitation workers not only endangers their health and human rights but also undermines broader public health efforts, especially in rapidly urbanising settings where risks of disease transmission are heightened. This neglect not only violates their rights as workers but also poses a significant public health threat to the wider community.^{3 12}

In Ethiopia, improving sanitation services is a national priority for enhancing public health and sustainable urban development.^{17 18} However, sanitation workers remain largely excluded from national labour regulations, leaving them without access to basic occupational health and safety (OHS) standards.^{8 19} This exclusion significantly compromises their well-being and undermines efforts to improve sanitation systems in the country.^{19 20} While sanitation services are crucial for reducing the burden of communicable diseases and improving the quality of life for urban populations, the precarious conditions faced by sanitation workers in Ethiopia further complicate the path toward achieving the SDGs.^{21 22} In cities like Addis Ababa, where the demand for sanitation services is growing due to rapid urbanisation, these workers often operate in environments that lack even basic safety measures, putting both their own health and the broader community at risk.^{23 24} Poor protection of sanitation workers has been linked to higher rates of enteric diseases and environmental contamination, increasing the risk of disease outbreaks in urban areas.³ Investing in the safety and well-being of sanitation workers serves both as a critical public health strategy and a fundamental human rights obligation.^{3 25}

Addressing these issues is essential not only from a legal perspective but also as a moral and humanitarian responsibility.^{7 26} Despite their indispensable contributions to public health and urban development, sanitation workers remain among the most underserved and under-researched labour sectors in Ethiopia. There is an urgent need to identify the barriers limiting their access to safe working conditions and to initiate multisectoral dialogue to ensure shared responsibility for their protection. Strengthening worker safety requires practical interventions, evidence-informed policies and inclusive planning that recognise the central role of sanitation workers in public health systems. Failure to address these challenges undermines the goal of achieving equitable and sustainable development, as envisioned in the SDGs, leaving a significant portion of the workforce vulnerable.²⁷ Evidence-based policy reform, capacity building and integration of sanitation worker health into national health strategies are critical to closing this equity gap and safeguarding the well-being of this essential workforce.

This study provides the first comprehensive assessment of access to safe sanitation services and PPE utilisation among sanitation workers in Ethiopia, a workforce critical to public health yet largely under-researched. By examining structural, behavioural and psychosocial factors influencing PPE use, including household size, workplace availability, supervision, cues to action, self-efficacy and perceived barriers, the study generates actionable, context-specific evidence to guide targeted interventions, occupational safety programmes and policy reforms. The findings establish a robust baseline for evaluating future interventions, ensuring that programmes are evidence-based and responsive to workers’ actual needs. Beyond practical applications, the research highlights the ethical, social and public health imperative of safeguarding sanitation workers, whose labour underpins hygiene, disease prevention and resilient health systems. By identifying critical gaps in protection and safety, the study provides essential insights for policymakers, public health institutions and development partners to enhance worker resilience, health and dignity, making their protection a national and strategic priority.

Project scope and objectives

This study was undertaken to achieve the following objectives:

1. To assess the level of access to safe sanitation services and PPE utilisation among sanitation workers.

2. To identify barriers that hinder sanitation workers from obtaining essential OHS protections.

3. To determine the factors associated with PPE utilisation among sanitation workers.

Methods

Project approach

To achieve the study objectives, the project was implemented in collaboration with existing institutional partners, including the Addis Ababa Water and Sewerage Authority (AAWSA) and other key stakeholders. These partnerships facilitated participant recruitment, stakeholder engagement and the dissemination of findings. Study findings were presented during a stakeholder workshop that engaged representatives from government agencies, sanitation service providers, labour organisations, academia and sanitation workers. The workshop provided a platform for discussing the implications of the findings and identifying practical strategies to strengthen OHS, improve access to safe sanitation services and support the social and economic inclusion of sanitation workers.

Study design and period

A workplace-based, cross-sectional concurrent mixed-methods study was conducted between November 2021 and June 2022. The quantitative component comprised a cross-sectional survey to assess sanitation workers’ access to safe sanitation services, PPE utilisation and associated factors. The qualitative component employed a descriptive qualitative approach using key informant interviews (KIIs) and focus group discussions (FGDs) to explore participants’ experiences, perceptions and contextual factors and to explain and contextualise the quantitative findings. The quantitative and qualitative findings were integrated during the interpretation stage to strengthen the validity and comprehensiveness of the study findings and to inform policy and practice.

Qualitative approach and research paradigm

The qualitative component employed a descriptive qualitative approach using KIIs and FGDs. This approach was selected because it enables an in-depth understanding of participants’ experiences, perceptions and contextual factors and is well-suited to explaining and contextualising the quantitative findings within an concurrent mixed-methods design rather than generating new theory. The qualitative inquiry was conducted within a postpositivist research paradigm, recognising that participants’ perspectives provide valuable but context-dependent insights into OHS practices. The integration of qualitative and quantitative evidence through methodological triangulation strengthened the credibility, completeness and interpretation of the study findings.

Study area

The study was conducted in Addis Ababa, the capital city of Ethiopia, which hosts a large workforce of public and private sanitation workers. Administratively, the city is divided into 11 sub-cities and 120 districts, with a population exceeding 4 million. According to the AAWSA, approximately 4521 sanitation workers were employed across the city during the study period.

The quantitative component was conducted across all sub-cities and selected districts to ensure a broad representation of sanitation workers. The qualitative component was undertaken in purposively selected sub-cities and workplaces representing government-employed, private and informal sanitation workers, where KIIs and FGDs were conducted. This urban context, characterised by rapid population growth and increasing sanitation demands, provided a relevant setting for examining both the prevalence and the contextual determinants of OHS among sanitation workers.

Study population and sample

The source population comprised all sanitation workers employed across the sub-cities of Addis Ababa. For this study, sanitation workers were defined as individuals engaged in any stage of the sanitation service chain, including toilet cleaning, pit latrine and septic tank emptying, sewer cleaning, waste collection and transport (manual or mechanical), wastewater treatment and final disposal, irrespective of whether they were employed in the public sector, private sector or informal sector.

For the quantitative component, the study population included sanitation workers who had at least 1 year of work experience and voluntarily consented to participate. Workers were excluded if they were critically ill, on maternity or sabbatical leave or unable to provide informed consent at the time of data collection.

For the qualitative component, participants were purposively selected sanitation workers, supervisors and relevant institutional stakeholders with direct experience in sanitation service delivery and occupational safety. Eligible qualitative participants had at least 1 year of experience and were able to provide informed consent. KIIs and FGDs were conducted to capture diverse perspectives from sanitation workers employed in the public, private and informal sectors as well as across different work settings.

Sample size determination and sampling procedure

To ensure adequate statistical power, the sample size for the quantitative component was determined using a single population proportion formula,²⁸ assuming a 50% proportion of sanitation workers with good health risk mitigation practices. This conservative estimate was used due to the absence of prior studies on the topic. A 95% CI and a 5% margin of error were applied. The formula used was:

$${\displaystyle {\displaystyle n=(Z\alpha /2)2[p(1-p)]/d2}}$$

where n=required sample size; Z=1.96 (standard normal value at 95% confidence level); p=estimated proportion (0.50) and d=margin of error (0.05).

This calculation yielded an initial sample size of 384. After adjusting for a 10% non-response rate and applying a design effect of 2 to account for cluster sampling variability, the final sample size was determined to be 844 sanitation workers.

For the qualitative component, sample size was guided by the concept of information power, which considers the study aim, sample specificity, use of established theory, quality of dialogue and analytic strategy to determine adequacy rather than numerical representativeness.^{29 30} Given the focused study aim, high specificity of participants, use of behavioural and occupational safety frameworks and in-depth data collection and analysis, a relatively small but information-rich sample was sufficient.

Accordingly, the qualitative component included a total of 77 participants, comprising 12 KIIs with sanitation supervisors, managers and institutional stakeholders, and 6 FGDs with sanitation workers. Data collection continued until thematic saturation was achieved, with no new concepts emerging in later interviews and discussions, indicating adequate information power for the study.

Quantitative data collection and variable measurement

Quantitative data were collected using a structured interviewer-administered questionnaire developed from the literature and adapted from previously validated instruments to align with the study objectives.^31–35 The questionnaire was reviewed by the research team for content validity, translated into Amharic by language experts and administered through face-to-face interviews conducted by trained data collectors at participants’ workplaces. The complete questionnaire is provided in online supplemental file 1.

The questionnaire comprised five sections assessing: (1) sociodemographic and occupational characteristics; (2) access to safe sanitation services; (3) behavioural characteristics (including cigarette smoking, alcohol consumption, khat chewing and sleep problems); (4) PPE use and reasons for non-use and (5) behavioural determinants of PPE use based on the Health Belief Model (HBM).

Access to safe sanitation services was assessed using 15 binary (Yes/No) items covering OHS training, first aid services, OHS committees, PPE availability and compensation for work-related injuries and illnesses.^{7 36–39} Participants scoring at or above the sample mean were classified as having adequate access to safe sanitation services. PPE utilisation was defined as the self-reported use of at least one task-appropriate PPE item (eg, gloves, safety boots, masks, goggles or helmets) while performing sanitation work during the previous month.^{31 34 35}

Behavioural determinants of PPE use were assessed using a 39-item HBM questionnaire adapted from previously validated instruments.^31–33 The questionnaire assessed six constructs: perceived susceptibility (6 items), perceived severity (7 items), perceived benefits (4 items), perceived barriers (8 items), cues to action (9 items) and self-efficacy (5 items). All items were measured using a 5-point Likert scale ranging from 1 (strongly disagree) to 5 (strongly agree). Negatively worded items were reverse-coded before analysis. Mean scores were calculated for each construct, and participants scoring at or above the sample mean were classified as having high perceived susceptibility, perceived severity, perceived benefits, cues to action and self-efficacy, whereas those scoring below the mean were classified as low. For perceived barriers, higher scores indicated greater perceived barriers to PPE use.

Body mass index (BMI) was classified as underweight (<18.5 kg/m²), normal (18.5–24.9 kg/m²) or overweight/obese (≥25.0 kg/m²).⁴⁰ Cigarette smoking was defined as smoking at least one cigarette per day,^{41 42} alcohol intake as consuming alcohol at least two times per week⁴¹ and khat chewing as chewing khat at least three times per week for a minimum of 12 months.^{42 43}

Qualitative data collection tools and procedures

Qualitative data were collected through KIIs and FGDs to explore barriers and facilitators related to access to safe sanitation services and the utilisation of PPE among sanitation workers. Semistructured interview guides were developed through a review of the literature and refined by the research team to align with the study objectives. The guides were prepared in English, translated into Amharic by experienced translators and pilot-tested before data collection. The complete KII and FGD guides are provided in online supplemental files 2 and 3, respectively.

12 purposively selected key informants from organisations responsible for sanitation, occupational health and labour regulation participated in the KIIs. Participants were selected based on their expertise and experience in sanitation services and occupational health. They included representatives from the Addis Ababa Health Bureau (n=2), Ethiopian Ministry of Health (n=2), AAWSA (n=4), Addis Ababa City Administration (n=1), Ethiopian Ministry of Labour and Skills (n=1) and Addis Ababa Municipality (n=2). Interviews were conducted by trained qualitative researchers in private settings at participants’ workplaces, audio-recorded with written informed consent and lasted approximately 40 min.

Six FGDs involving 65 participants were conducted to obtain broader stakeholder perspectives. Participants were purposively selected during a stakeholder workshop to ensure representation from sanitation workers, sanitation worker representatives, government agencies, labour organisations, academic institutions and water and sanitation authorities. Each FGD included 8–12 participants and lasted approximately 60–120 min. Discussions were facilitated by trained moderators using semistructured interview guides, supported by note-takers and audio-recorded with written informed consent. Data collection continued until thematic saturation was achieved.

Audio recordings were transcribed verbatim and translated into English where necessary before analysis. Qualitative data were analysed thematically to identify recurring patterns, themes and explanations for the quantitative findings. Trustworthiness was enhanced through triangulation of data sources (KIIs and FGDs), inclusion of participants with diverse perspectives, use of a consistent semistructured interview guide, maintenance of an audit trail and reflexive memos and grounding interpretations in verbatim quotations. Detailed descriptions of the study context, participants and settings were provided to support the transferability of the findings.

Training

The study team conducted a 2-day training for 22 research assistants, comprising 6 supervisors and 16 data collectors, all of whom were proficient in Afan Oromo and Amharic. The training aimed to equip the team with the necessary skills and knowledge to effectively recruit participants and administer the survey. The training sessions focused on key areas such as research methods, ethical considerations, survey administration procedures and maintaining confidentiality.

In addition to the foundational training, the research assistants worked closely with the team to review and refine each survey item, ensuring clarity and improving the relevance of the measures to the local context and population. This collaborative process helped tailor the survey to better address the unique needs and experiences of the target group, enhancing the overall quality and applicability of the data collection process.

Data quality assurance

To ensure data quality, substantial emphasis was placed on the careful design and validation of data collection tools. The questionnaire was initially developed in English and reviewed by all project team members, whose feedback was incorporated to improve clarity, relevance and content validity. The tool was then translated into Amharic and back-translated into English by bilingual project members with support from language experts to ensure semantic consistency. A pre-test was conducted on 5% of the sample in Gondar City 1 week before data collection to assess clarity, appropriateness and internal consistency. Based on pre-test findings, revisions were made, including refining wording, resolving ambiguities and reducing the number of questions without compromising the intended measurements. During data collection, supervisors and project members provided close oversight, and any emerging issues were promptly addressed through on-site discussions to maintain consistency and reliability.

Trustworthiness of the qualitative data

Trustworthiness of the qualitative data was ensured using established criteria of credibility, dependability, confirmability and transferability. Credibility was enhanced through triangulation of data sources (KIIs and FGDs), inclusion of participants from diverse employment types and prolonged engagement during data collection. Dependability was supported by the use of a consistent semistructured interview guide and a transparent, systematic analytic process. Confirmability was strengthened through reflexive memo-writing, maintenance of an audit trail and grounding interpretations in verbatim participant quotations. Transferability was facilitated by providing detailed descriptions of the study context, participants and settings, enabling readers to assess the applicability of the findings to similar contexts.

Quantitative data analysis

The collected data were checked for completeness before analysis. Accurate and complete records were coded, labelled, categorised and entered into EpiData V.4.6 software. The cleaned dataset was subsequently exported to STATA V.14 for statistical analysis. Descriptive statistics, including frequencies, percentages, means and 95% CIs, were computed and presented using tables, figures and narrative summaries. The internal consistency of all measurement instruments was assessed using Cronbach’s alpha. The items evaluating access to safe sanitation services demonstrated good reliability (α=0.76), while the measure of PPE utilisation showed excellent internal consistency (α=0.87). The six constructs derived from the HBM also demonstrated acceptable to excellent reliability, with Cronbach’s alpha values of 0.81 for perceived susceptibility, 0.84 for perceived severity, 0.79 for perceived benefits, 0.83 for perceived barriers, 0.88 for cues to action and 0.86 for self-efficacy. All reliability coefficients exceeded the recommended threshold of 0.70, indicating good internal consistency.

Before conducting inferential analyses, the assumptions underlying logistic regression were assessed. The distribution of continuous variables was evaluated for normality, potential outliers were examined and multicollinearity was assessed using the Variance Inflation Factor (VIF). All predictors yielded VIF values below 5, indicating no evidence of problematic multicollinearity and supporting the validity of the model specification.⁴⁴

Candidate independent variables were selected a priori based on epidemiological relevance, evidence from previous literature and the conceptual framework of the HBM. Candidate variables included sociodemographic characteristics (eg, age, sex, educational level, family size and years of sanitation work), occupational characteristics (eg, employment type, type of sanitation activity, occupational safety and health training, PPE training, PPE availability and workplace supervision), behavioural characteristics and HBM constructs. Each variable was initially examined using bivariable binary logistic regression, and variables with a p value <0.20 were entered into the multivariable logistic regression model to control for potential confounding.

Statistical significance in the multivariable model was determined at p value <0.05, and the strength of associations was reported using adjusted ORs (AORs) with 95% CIs. Model goodness-of-fit was assessed using the Hosmer-Lemeshow test, which indicated an acceptable fit (p value >0.05), confirming that the final model adequately represented the observed data.

Qualitative data analysis

Interview data were recorded in Amharic and subsequently manually transcribed from the audio files. Following transcription, the data were translated into English to ensure clarity and accuracy in interpretation. The translated transcripts were then uploaded into OpenCode V.4.03, a robust qualitative data analysis software, for coding and organising the qualitative data. A coding team convened weekly to identify emerging themes from the dataset and to ensure that codes were applied appropriately across various contexts.

To maintain analytic rigour and consistency, the coding process was conducted in two stages. In the first stage, a preliminary codebook was developed from close reading of the initial interview and focus group transcripts, particularly the most information-rich interviews, rather than relying solely on field notes. Codes were generated deductively based on the study objectives and conceptual frameworks, and inductively as new concepts emerged from participants’ narratives during early transcript analysis. As coding progressed, the codebook was iteratively refined to incorporate emerging insights, resulting in a final codebook comprising 195 codes, which was applied during the second stage of coding across all transcripts. To enhance reliability, two independent researchers coded the data separately and then compared coding outputs. Coding discrepancies were discussed and resolved through consensus, leading to a shared interpretation of codes and themes. A high level of agreement was achieved through this iterative consensus process, and consistency was further supported by regular team discussions and refinement of code definitions throughout the analysis. On completion of the coding process, qualitative tools within OpenCode were used to identify emerging themes, with a particular focus on code co-occurrences related to barriers in accessing occupational exposure prevention practices. These co-occurrences were systematically analysed to determine the key themes present in the data. All excerpts related to these barriers were carefully categorised and organised into major themes, which were then compiled in an Excel spreadsheet for further analysis.

The project team meticulously refined the themes, identified commonalities and developed a comprehensive write-up of the findings. This thematic analysis was integrated with the quantitative results of the study, providing a holistic view of the challenges and insights related to occupational exposure prevention practices. The integration of both qualitative and quantitative findings strengthened the study’s conclusions and offered a more nuanced understanding of the factors affecting sanitation workers’ health and safety practices.

Patient and public involvement statement

There was no participant, patient or public involvement in the design and conduct of the study. The study results were shared with the participants.

Results

Sociodemographic characteristics of study participants

A total of 821 sanitation workers participated in the study, yielding a response rate of 97.27%. The majority of participants were male (n=596, 72.6%). The age of respondents ranged from 19 to 65 years, with a mean age of 35.81 years (SD ±8.60). One-third (45.3%) of participants had completed primary education, while 27.0% (n=221) had completed secondary education. More than half of the respondents (n=514, 62.6%) reported being married. The mean duration of work experience among participants was 6 years (SD ±4.56), with 44.3% (n=364) reporting five or more years of experience. The average monthly salary was 5337 Ethiopian Birr (ETB), with 29.1% (n=239) earning <2500 ETB. In terms of work shifts, the vast majority (n=729, 88.8%) worked exclusively during the day, while 11.0% (n=90) reported alternating between day and night shifts. A small proportion (n=2, 0.2%) worked exclusively at night (table 1).

Table 1. Sociodemographic characteristics of sanitation workers in Addis Ababa, the capital and largest city of Ethiopia (N=821).

Variables	Frequency (n)	Percent (%)
Sex
Male	596	72.6
Female	225	27.4
Age (in years)
19–29	203	24.7
30–39	370	45.1
≥40	248	30.2
Religion
Orthodox	531	64.7
Muslim	57	6.9
Protestant	230	28.0
Other*	3	0.4
Ethnicity
Oromo	317	38.6
Amhara	255	31.1
Tigre	53	6.5
Gurage	7	0.9
Silte	8	1.0
Hadiya	7	0.9
Other†	174	21.2
Educational level
Unable to read and write	11	1.3
Primary school completed	372	45.3
Secondary school completed	221	27.0
Diploma and above	217	26.4
Marital status
Single	270	32.9
Married	514	62.6
Divorced	22	2.7
Widowed	15	1.8
Years in sanitation work
<5 years	457	55.7
≥5 years	364	44.3
Working days per week
<5 days	141	17.2
≥5 days	680	82.8
Working hours per day
≤8 hours	565	68.8
>8 hours	256	31.2
Family size
≤2	239	25.9
3–4	389	47.4
≥5	219	26.7
Monthly income (ETB)
<2500	239	29.1
2500–4000	211	25.7
4001–7500	180	21.9
>7500	191	23.3
Work shift
Only night shift	2	0.2
Only day shift	729	88.8
Work both shifts interchangeably	90	11.0

^*Catholic and Adventist.

^†Not specified.

ETB, Ethiopian Birr (1US$=54.35 ETB).

Occupational hazards and inadequate compensation

Quantitative findings indicated that a large proportion of sanitation workers received low wages despite engaging in physically demanding and hazardous work. Many reported challenges in meeting basic living expenses and covering health-related costs.

Qualitative findings elaborate on this pattern through two categories: (1) low wages relative to occupational exposure and (2) limited compensation for work-related health risks.

One participant stated:

The payment sanitation workers receive does not match the level of risk they face every day. They handle waste and unsafe materials, but their income is too low to cover treatment costs when they become sick or injured. (FGD participant)

Participants also reported that compensation mechanisms for occupational illness or injury were limited or absent, and that sanitation workers were often expected to continue working despite health concerns. These accounts were consistently reported across interviews and group discussions.

Conditions of employment

Quantitative findings showed that sanitation workers were employed under diverse arrangements. Of the 821 participants, 422 (51.4%) were employed in the public sector, 152 (18.5%) worked for private employers without formal registration and 141 (17.2%) reported being self-employed under formally registered arrangements (figure 1). The remaining participants reported short-term or daily paid work.

Figure 1. Sanitation workers’ condition of employment.

Qualitative findings described these employment conditions under the theme ‘Informal and unstable employment arrangements’, with two categories: (1) temporary and daily paid work and (2) lack of formal registration and supervision.

Participants reported that many sanitation workers were engaged on a temporary or daily payment basis, particularly outside permanent government positions. Employment arrangements were described as unstable, with irregular income and unclear terms of engagement. Several participants stated that some workers were not formally registered, licensed, or consistently supervised, making them difficult to track within official systems.

One participant reported:

There are sanitation workers who are formally employed, but others are not registered or licensed. Some are hired temporarily and are paid daily, and they are not regularly supervised. (KII participant)

Across interviews and group discussions, participants consistently described the coexistence of permanent, temporary and informal work arrangements within the sanitation sector.

Nature of employment and dignity of workers

Quantitative findings showed that among the 821 sanitation workers surveyed, 476 (58.0%) reported permanent employment, while 263 (32.0%) and 82 (10.0%) were employed on a temporary and casual basis, respectively (see online supplemental figure S1).

Qualitative findings described experiences related to dignity and social perception under the theme ‘Employment status and social stigma’ with two categories: (1) perceived dignity within the workforce and (2) stigma and discrimination from the community.

Participants consistently reported that stigma or abuse was uncommon among sanitation workers themselves, with workers describing mutual respect and solidarity within the workforce. In contrast, stigma was frequently reported as originating from the wider community, where sanitation work was associated with sewage, waste and unpleasant odours. Participants described differential treatment in public spaces and social interactions once their occupation was known.

One participant stated:

Among the workers, there is no stigma. But in the community, people treat sanitation workers differently because they deal with sewage. Some avoid greeting them or sitting near them when they know what work they do. (KII participant)

Another participant added a related observation regarding public behaviour:

When sanitation workers are present in public places, some people cover their noses or step away. This happens even though the workers themselves maintain hygiene. (FGD participant)

Across interviews and FGDs, participants consistently described social avoidance and reduced social acceptance associated with sanitation work.

Types of work performed by sanitation workers

Quantitative findings indicated that sanitation workers were engaged in a wide range of sanitation-related tasks, including septic tank emptying, toilet cleaning, drain and sewer cleaning, operation and maintenance of pumping stations, waste transportation, final disposal and waste reuse activities. Among the 821 respondents, 207 (25.2%) were primarily involved in toilet cleaning, 158 (19.2%) in sewer cleaning, 146 (17.8%) in septic tank emptying and 156 (19.0%) worked at final disposal sites (figure 2).

Figure 2. Types of work sanitation work performed by sanitation workers.

Qualitative findings described the nature of these tasks under the theme ‘Work activities involving hazardous environments’. Participants reported that sanitation work frequently involved exposure to strong odours and gases, particularly during sewer, septic tank and final disposal activities, and that these exposures were more intense at waste treatment and disposal sites.

One participant described this experience as follows:

When workers enter sewer lines or work at the disposal sites, the smell is very strong. Gases like hydrogen sulfide and methane are present, and workers are exposed to them for long periods, especially at the final treatment plants. (KII participant)

Across interviews and discussions, participants consistently described sanitation tasks as physically demanding and carried out in environments characterised by close contact with waste and unpleasant working conditions, with the level of exposure varying by type of work performed.

Behavioural characteristics of study participants

According to the WHO classification of BMI, 61 (7.4%) of the sanitation workers were classified as underweight, 634 (77.2%) had a normal BMI and 126 (15.4%) were categorised as overweight or obese. Regarding substance use, 150 (18.3%) of participants reported alcohol consumption, 47 (5.7%) reported chewing khat and 35 (4.3%) reported smoking cigarettes (see online supplemental table S1).

Access to safe sanitation services

Quantitative findings showed that 315 (38.2%) sanitation workers reported having good access to safe sanitation services (95% CI 34.9 to 41.6). A majority, 558 (68.0%), reported that they had not received training related to safe sanitation work. In addition, 579 (70.5%) stated that no workplace guidelines or regulations on safe working practices were available. Pre-employment medical examinations were absent for 510 (62.1%) workers, and 743 (90.5%) reported that periodic medical check-ups were not conducted. Most respondents also reported a lack of access to first aid services (684; 83.3%) and absence of workplace safety committees (748; 91.1%). Furthermore, 544 (33.7%) indicated that compensation services for work-related injuries were unavailable and 588 (71.6%) reported that employers did not provide vaccinations against tetanus, tuberculosis, hepatitis or COVID-19 (table 2).

Table 2. Access to safe sanitation service among sanitation workers in Addis Ababa, Ethiopia (N=821).

Variables	Frequency	Percent
Training on the health and safety of sanitation work
Yes	263	32.0
No	558	68.0
Guidelines and regulations are available on safe work procedures
Yes	242	29.5
No	579	70.5
Pre-employment medical examination available
Yes	311	37.9
No	510	62.1
Periodic/or special medical examinations are available
Yes	78	9.5
No	743	90.5
First aid available
Yes	137	16.7
No	684	83.3
Health and safety committee available
Yes	73	8.9
No	748	91.1
Are health and safety supervised by the government?
Yes	92	11.2
No	729	88.8
Availability of PPE
Yes	301	36.7
No	520	63.3
Training on the use of PPE is available
Yes	165	20.1
No	656	79.9
Compensation for injuries and occupational diseases is available
Yes	277	33.7
No	544	66.3
Are there treatment services for injuries and illnesses provided by the employer
Yes	325	39.6
No	496	60.4
Welfare facilities such as a toilet, a changing room, washing water and soap, and separate eating and drinking rooms at work are available
Yes	468	57.0
No	353	43.0
Vaccination offered by the employer for diseases such as tetanus, TB, hepatitis and COVID-19
Yes	233	28.4
No	588	71.6
Deworming service available during the transition from unsafe to safe practices
Yes	178	21.7
No	643	78.3
Overall access to safe sanitation services
Good access	314	38.2
Poor access	507	61.8

PPE, personal protective equipment; TB, tuberculosis.

Qualitative findings described these conditions under the theme ‘Limited access to occupational health and safety services’, with reported gaps across training, medical surveillance, safety infrastructure and protective provisions. Participants stated that although policies and regulations related to sanitation worker safety existed, they were not consistently implemented at the workplace level. Reports included a lack of training opportunities, an absence of PPE provision and limited monitoring of safety practices.

One participant described this as follows:

There are laws and policies that are meant to protect sanitation workers, but in practice, they are not applied. Many workers do not receive basic protective equipment like gloves, safety shoes or clothing, and there is little follow-up to check whether safety rules are being followed. (KII participant)

Across interviews and discussions, participants consistently reported limited availability of organised OHS services within sanitation workplaces.

Sources of information on safe sanitation work

Among the sanitation workers surveyed, social media platforms, particularly Facebook and Telegram, were identified as the most common sources of information on safe sanitation practices, reported by 32.9% of participants. This was followed by information received from colleagues or employers at the workplace (24.9%) and television broadcasts (21.8%) (figure 3). These findings underscore the increasing role of digital platforms in disseminating OHS information, while also highlighting the continuing relevance of peer communication and traditional media.

Figure 3. Source of information on safe sanitation work among sanitation workers.

Accessibility and utilisation of PPE

Quantitative findings showed that 446 (54.3%) sanitation workers reported using at least one type of PPE during their work (95% CI 50.8 to 57.8). Face masks were the most commonly used item (53.0%), followed by sanitary gloves (49.9%). In contrast, safety aprons (0.7%) and safety glasses (3.8%) were rarely used. Safety shoes were used by 40.7% of workers, while 37.6% reported wearing protective clothing during sanitation-related tasks (figure 4). Marked differences were observed by employment type. Government-employed sanitation workers reported the highest PPE utilisation (59.9%), whereas self-employed workers reported the lowest use (12.7%) (figure 5).

Figure 4. Utilisation of personal protective equipment (PPE) among sanitation workers.

Figure 5. Distribution of the use of personal protective equipment (PPE) in employment conditions among sanitation workers.

Qualitative findings described PPE access and use under the theme ‘Irregular provision and inconsistent use of PPE’, with two categories: (1) delayed and insufficient PPE supply and (2) lack of routine replacement and monitoring.

Participants described prolonged delays in PPE procurement and distribution, resulting in workers performing sanitation tasks without adequate protection. PPE distribution was often reported to occur as a one-time event, with no clear plan for replacement or follow-up. Participants also described limited attention to PPE use after distribution.

One participant explained this in detail:

The process of purchasing PPE takes a very long time. By the time the equipment arrives, workers may have already spent months or even years working without any protection. When PPE is finally distributed, it is often given only once, in bulk, and there is no system to check whether it is still usable or needs replacement. As a result, many workers continue their daily activities without proper protection, especially when the equipment becomes worn out or damaged. (FGD participant)

Across interviews and FGDs, participants consistently reported challenges related to the availability, continuity and management of PPE in sanitation workplaces.

Institutional and behavioural factors affecting the utilisation of PPE

Among the surveyed sanitation workers, 520 (63.3%) reported that PPE was not available in their workplace. Of those who had access to PPE (n=301), approximately half (n=152, 50.5%) stated that the equipment was provided by the government, whereas 71 (23.6%) reported having to purchase PPE using their own resources. In terms of training, a significant majority of respondents (676; 82.3%) had not received instruction on the proper use of PPE (see online supplemental table S2). This lack of training likely contributed to low compliance with PPE usage protocols. Respondents identified several key barriers to wearing PPE. The most frequently cited reason was the unavailability of PPE (59.2%), followed by lack of awareness regarding its effectiveness (20.0%). Additional reported barriers included discomfort while wearing PPE (12.3%), and the perception that PPE use slows down work performance (8.5%) (see online supplemental figure S2).

Factors associated with PPE utilisation

A multivariable logistic regression analysis was conducted to identify factors associated with PPE utilisation among sanitation workers. In adjusted multivariable logistic regression analysis, family size, availability of PPE at the workplace, supervision of PPE use, perceived barriers, cues to action and self-efficacy were significantly associated with PPE utilisation. Although several variables showed strong crude associations with PPE utilisation, some associations were attenuated after multivariable adjustment, reflecting shared variance among related occupational and behavioural factors.

Workers from households with three to four members were 1.58 times more likely to use PPE than those from households with two or fewer members (AOR=1.58; 95% CI 1.02 to 2.47; p=0.041). Similarly, workers from households of five or more people were 2.44 times more likely to use PPE (AOR=2.44; 95% CI 1.43 to 4.17; p=0.001). Workers with access to PPE in the workplace were six times more likely to use it than those without access (AOR=6.10; 95% CI 4.01 to 9.31; p<0.001). Active supervision of PPE use markedly increased its utilisation: supervised workers were nearly eight times more likely to use PPE than unsupervised workers (AOR=7.95; 95% CI 3.22 to 19.66; p<0.001). Perceived barriers to PPE use reduced the likelihood of use by 52% (AOR=0.48; 95% CI 0.33 to 0.71; p<0.001). Conversely, high cues to action regarding PPE use increased the likelihood of using PPE by 69% (AOR=1.69; 95% CI 1.14 to 2.52; p=0.010). High self-efficacy regarding PPE use was also associated with an 81% higher likelihood of using PPE (AOR=1.81; 95% CI 1.18 to 2.77; p=0.006), as shown in table 3.

Table 3. Logistic regression analysis for factors associated with PPE utilisation among Sanitation Workers in Ethiopia.

Variables	Utilisation of PPE		COR with 95% CI	P value	AOR with 95% CI	P value
	Yes	No
Age
19-29	97	106	1		1
30-39	181	189	1.05 (0.74 to 1.47)	0.795	0.82 (0.53 to 1.31)	0.418
40-49	130	54	2.63 (1.73 to 4.00)	0.001	1.48 (0.85 to 2.58)	0.161
50-65	38	26	1.59 (0.90 to 2.82)	0.107	0.99 (0.48 to 2.08)	0.999
Family size
≤2	84	129	1		1
3-4	210	179	1.80 (1.28 to 2.53)	0.001	1.58 (1.02 to 2.47)	0.041
≥5	152	67	3.48 (2.34 to 5.19)	<0.001	2.44 (1.43 to 4.17)	0.001
OHS training
Yes	192	71	3.23 (2.35 to 4.45)	<0.001	1.02 (0.66 to 1.59)	0.921
No	254	304	1		1
Availability of PPE
Yes	256	45	9.88 (6.87 to 14.21)	<0.001	6.10 (4.01 to 9.31)	<0.001
No	190	330	1		1
Supervision on PPE use
Yes	105	7	16.19 (7.43 to 35.28)	<0.001	7.95 (3.22 to 19.66)	<0.001
No	341	368	1		1
Trained on how to use PPE
Yes	324	23	5.76 (3.60 to 9.22)	<0.001	0.79 (0.41 to 1.52)	0.480
No	122	352	1		1
Perceived susceptibility to occupational illness
High	222	271	0.38 (0.28 to 0.51)	<0.001	0.48 (0.30 to 1.77)	0.126
Low	224	104	1		1
Perceived severity of occupational illness
High	193	257	0.350 (0.26 to 0.47)	<0.001	0.72 (0.46 to 1.14)	0.169
Low	253	118	1		1
Perceived barrier to using PPE
High	173	249	0.32 (0.24 to 0.43)	<0.001	0.48 (0.33 to 0.71)	<0.001
Low	273	126	1		1
Cues to action for PPE use
High	304	195	1.98 (1.48 to 2.63)	<0.001	1.69 (1.14 to 2.52)	0.010
Low	142	180	1		1
Self-efficacy for PPE use
High	303	196	1.94 (1.45 to 2.57)	<0.001	1.81 (1.18 to 2.77)	0.006
Low	143	179	1		1

^*Significant at a p value <0.05 in multivariable logistic regression analysis, Hosmer and Lemeshow test p value= 0.1563

1, reference category; AOR, adjusted OR; COR, crude OR; OHS, occupational health and safety; PPE, personal protective equipment.

Discussion

This mixed-methods study identified substantial gaps in both access to safe sanitation services and PPE use among sanitation workers in Ethiopia. Only 38.2% (95% CI 34.9% to 41.6%) of participants reported adequate access to safe sanitation services, while 54.3% (95% CI 50.8% to 57.8%) reported using PPE. PPE use was significantly associated with organisational and behavioural factors, including workplace PPE availability, supervision, cues to action and self-efficacy, whereas perceived barriers were associated with lower odds of PPE use. Qualitative findings complemented these results by highlighting inconsistent PPE provision, limited occupational health training, weak regulatory enforcement, precarious employment conditions and low social recognition of sanitation work as major barriers to safe workplace practices. Together, these findings indicate that improving worker protection requires coordinated interventions that address both individual behavioural factors and systemic workplace conditions.

Access to safe sanitation services

This study revealed that sanitation workers in Ethiopia have limited access to safe sanitation services, consistent with evidence from other LMICs. Our findings are in agreement with those of a multicountry study conducted in Burkina Faso, Nigeria, Tanzania and Zambia, which reported similar OHS challenges among sanitation workers, including inadequate sanitation infrastructure, limited workplace facilities and persistent social stigma.⁷ In the absence of functional faecal sludge treatment plants, workers were often forced to dispose of faecal sludge in open environments, further compromising their dignity and increasing social discrimination.⁷

The qualitative findings provide important context for these quantitative results. Participants consistently described inadequate institutional support, limited occupational health infrastructure, inconsistent PPE provision and weak implementation of OHS regulations as major barriers to safe working conditions. Together, these findings suggest that limited access to safe sanitation services reflects broader organisational and systemic shortcomings rather than isolated workplace deficiencies.

Most sanitation workers in the present study had not received OHS training before commencing work. This finding is consistent with a study conducted in Bangladesh, where most solid waste and sanitation workers lacked infection prevention training and worked under unsafe conditions with limited institutional support.⁴⁵ Similarly, a study from the West Guji Zone of Oromia, Ethiopia, reported that 57.4% of solid waste workers had not received any occupational safety training before starting work.⁴⁶ In contrast, another Ethiopian study reported that 42.6% of solid waste collectors had received safety training before commencing their duties,⁴⁷ while more than 95% of construction workers in South Korea had received occupational safety training.⁴⁸ These differences may reflect variations in the implementation and enforcement of OHS regulations, resource allocation and institutional commitment across settings.

The qualitative findings further explained the low level of OHS training observed in the survey. Participants reported that training was rarely prioritised, particularly for temporary and informal sanitation workers, and when provided, it was often irregular and insufficient to prepare workers for occupational hazards. These findings suggest that limited institutional investment and weak implementation of occupational safety programmes may contribute to the low training coverage observed in this study.

Only 36.7% of sanitation workers reported adequate availability of PPE, substantially lower than the 70.4% reported among sanitation workers in Oromia, Ethiopia.⁴⁶ The qualitative findings provided important context for this difference, with participants describing frequent interruptions in PPE supply, delayed replacement of damaged equipment, inconsistent distribution practices and inadequate employer commitment to worker safety. These findings suggest that limited PPE availability may reflect organisational, procurement and resource constraints rather than individual worker preferences.

Inadequate funding, weak enforcement of occupational safety legislation and ineffective procurement practices may also contribute to limited PPE availability. In many local government settings, PPE is procured in bulk primarily to minimise costs, often without considering workers’ individual needs. Consequently, sanitation workers may receive oversized or undersized PPE, reducing both comfort and protective effectiveness. Ill-fitting PPE has been shown to impair mobility, reduce compliance and increase the risk of occupational injury, particularly in high-risk occupations.⁴⁹ Furthermore, PPE is commonly designed using male anthropometric data, limiting its suitability for female sanitation workers and contributing to gender inequities in occupational health protection.⁵⁰

Beyond structural barriers, behavioural and informational factors may also influence PPE use. Many sanitation workers reported limited awareness of occupational safety standards and their workplace rights. Consistent with the quantitative findings on behavioural determinants, participants described limited access to occupational safety information, inadequate managerial support and poor workplace safety culture as barriers to consistent PPE use. Previous studies have similarly shown that inadequate occupational health training and limited access to social and economic support reduce workers’ risk perception and motivation to use protective equipment.⁴⁹ These findings highlight the need for comprehensive, rights-based occupational health strategies that combine sustained PPE provision, gender-responsive PPE design, effective regulatory enforcement and context-specific occupational safety education.

The absence of accessible occupational health services, including vaccination, routine health screening and emergency medical support, further increases occupational health risks among sanitation workers. Participants described frequent exposure to hazardous chemicals, biological agents and work-related illnesses, including tetanus, hepatitis A, typhoid, dermatological conditions and respiratory diseases. Similar barriers to accessing occupational health services have been reported among sanitation workers in other African countries.⁷ Improving access to preventive and curative occupational health services should therefore remain a public health priority for this vulnerable workforce.

Consistent with the quantitative findings, the qualitative data also highlighted substantial gaps in the implementation and enforcement of OHS policies. Despite the existence of national legislation intended to protect sanitation workers, participants reported weak regulatory oversight, inadequate employer accountability and inconsistent monitoring of workplace safety standards. A key informant from the Addis Ababa Health Office noted that employers frequently failed to provide essential PPE, while regulatory authorities lacked effective mechanisms to ensure compliance. Together, these findings suggest that strengthening regulatory enforcement, employer accountability and routine occupational health monitoring may improve worker protection and contribute to safer sanitation workplaces.

Utilisation of PPE

In this study, the overall utilisation of PPE among sanitation workers was 54.3% (95% CI 50.8% to 57.8%). Face masks and gloves were the most commonly used PPE, consistent with findings from Pakistan and Ethiopian factory workers, where these items were the most frequently used forms of protection.^{31 51} However, most participants (63.3%) reported that PPE was unavailable at their workplace, highlighting persistent deficiencies in workplace safety provisions.

The prevalence of PPE utilisation was comparable to that reported among waste handlers in Bule Hora, Ethiopia (55%)⁴⁶ and sanitation workers in southern Ethiopia (51%).⁵² However, it was lower than that reported among informal waste workers in Kathmandu (68%),⁵³ municipal solid waste workers in Chandigarh, India (74.5%)⁵⁴ and landfill waste workers in Nepal (60%).⁵⁵ These differences may reflect variations in OHS systems, employer compliance with safety regulations, the availability and consistent supply of PPE, access to occupational health training, workplace supervision and the proportion of workers employed in the formal sector. Although differences in the operational definition of PPE use across studies may have contributed to some variation, contextual and organisational factors are more likely to explain the higher PPE utilisation reported in these settings.

Conversely, PPE utilisation in the present study was higher than that reported among waste workers in Kampala (33%) and waste collectors in Addis Ababa (43.6%),^{56 57} as well as factory workers in Debre Berhan (41.7%).³¹ These differences may reflect variation in occupational hazards, workplace safety practices, regulatory oversight and supervision across occupations and settings. The inherently hazardous nature of sanitation work, including exposure to biological hazards and toxic gases, may also increase workers’ perceived need to use PPE.

The quantitative findings suggest that inadequate PPE availability and limited occupational safety training may contribute to suboptimal PPE use. Consistent with this, most participants (82.3%) reported that they had not received training on the correct use of PPE. The qualitative findings further explained these quantitative results by showing that gaps in PPE utilisation were driven largely by organisational weaknesses rather than individual unwillingness to use protective equipment. Participants described prolonged delays in PPE procurement and distribution, irregular replacement of worn or damaged equipment and inconsistent supervisory follow-up. Some workers also reported purchasing PPE at their own expense when employers failed to provide adequate equipment. Similar challenges have been documented in other low-resource settings, where sanitation workers frequently experience inconsistent PPE supply and weak occupational safety management systems.^{34 58} These findings suggest that strengthening organisational systems for PPE procurement, distribution and workforce training may be more effective than relying solely on individual behaviour change to improve PPE utilisation.

The qualitative findings further indicated that consistent PPE provision serves as an important indicator of organisational commitment to worker safety. Participants reported that irregular PPE supply, inadequate supervision, poor fit, discomfort and limited training undermined confidence in workplace safety and contributed to inconsistent PPE use. These findings are consistent with previous research demonstrating that PPE utilisation is shaped by interacting organisational and behavioural factors rather than knowledge alone.⁵⁹ Together, the quantitative and qualitative findings suggest that improving PPE utilisation requires integrated interventions that strengthen PPE procurement and replacement systems, workplace supervision, occupational health training and enforcement of OHS standards.⁵⁸

Factors associated with PPE utilisation

Household size was significantly associated with PPE utilisation, with workers from larger households more likely to use PPE than those from smaller households. This association may reflect greater risk perception and a stronger sense of responsibility to protect household health and financial security, as work-related illness or injury could adversely affect family well-being. Similar findings have been reported in occupational health and behavioural studies, where individuals with greater family responsibilities are more likely to adopt preventive health behaviours because of increased risk aversion and economic considerations.^60–62

Workplace PPE availability was independently associated with PPE utilisation. Consistent access to functional, task-appropriate and well-fitting PPE is fundamental to safe work practices. The qualitative findings further indicated that reliable PPE provision reflects organisational commitment to worker safety, whereas irregular supply, delayed replacement and poor-quality equipment undermine consistent PPE use. Similar organisational barriers have been reported in other low- and middle-income settings, where inadequate procurement systems and inconsistent PPE supply limit adherence to recommended protective practices.^{3 7 10 38 39 63 64}

Regular workplace supervision was another significant predictor of PPE utilisation. Supervision may promote PPE use by reinforcing safety expectations, providing immediate feedback, correcting unsafe practices and strengthening organisational safety culture. The qualitative findings complemented this association by showing that workers perceived active supervisory engagement as an indication that their health and safety were valued, thereby encouraging consistent adherence to PPE use. Similar findings have been reported in previous occupational health studies.^{37 64–66}

Although OHS training and PPE-specific training were significantly associated with PPE utilisation in the bivariable analysis, these associations were attenuated after multivariable adjustment. This attenuation likely reflects shared variance among workplace characteristics and HBM constructs rather than problematic multicollinearity. Workers who received training were also more likely to report adequate PPE availability, regular supervision and higher levels of perceived susceptibility, perceived severity and self-efficacy. Given the cross-sectional design, these relationships should be interpreted as associations rather than causal pathways. Together, the quantitative and qualitative findings suggest that training alone may be insufficient to improve PPE utilisation without concurrent organisational support, including reliable PPE provision, effective supervision and a positive workplace safety culture.

Consistent with the HBM, perceived barriers were negatively associated with PPE utilisation, whereas cues to action and self-efficacy were positively associated with PPE use. Workers reporting greater perceived barriers were less likely to use PPE, while those with stronger cues to action and higher self-efficacy demonstrated greater adherence to protective practices. The qualitative findings provided further insight by showing that discomfort, poor PPE fit, inconsistent availability, limited managerial support and inadequate replacement of damaged equipment contributed to these perceived barriers. Similarly, regular reminders, supervisory encouragement and supportive workplace environments appeared to reinforce cues to action and workers’ confidence in using PPE correctly. Together, these findings suggest that behavioural determinants of PPE use are closely intertwined with organisational conditions, highlighting the importance of integrated interventions that address both workplace systems and individual behavioural factors.^{7 9 33 34 67–70}

Strengths and limitations

This study employed a concurrent mixed-methods design, a large sample and a multistage random sampling approach, strengthening methodological rigour and facilitating the integration of quantitative and qualitative findings. Several limitations should be acknowledged. Self-reported data and interviewer-administered questionnaires may have introduced recall, social desirability and interviewer bias. Although the HBM provided a useful framework for examining behavioural determinants of PPE use, it may not fully capture broader organisational and contextual influences. In addition, dichotomisation of HBM constructs using the sample mean may have resulted in misclassification, reduced variability and statistical power and limited comparability with other studies. PPE utilisation was defined as the use of at least one task-appropriate PPE item, which may overestimate comprehensive protection. Finally, the cross-sectional design precludes causal inference; therefore, the observed relationships should be interpreted as associations rather than causal effects, and the findings may not be generalisable beyond similar urban settings.

Recommendations and policy implications for practice

Improving access to safe sanitation services and strengthening PPE utilisation must be positioned as core priorities within Ethiopia’s sanitation and occupational health agendas. This study revealed substantial gaps in essential safety provisions, including inconsistent PPE supply, limited health and safety training and weak institutional support, highlighting the urgent need for national and regional policies that explicitly mandate employers to provide appropriate PPE, first aid services, social protection benefits and functional OHS structures. Embedding these requirements in regulatory frameworks, service contracts and monitoring systems will help standardise protections and ensure accountability across both public and private sanitation operations.

Several modifiable factors were significantly associated with PPE utilisation and provided clear targets for intervention. Regular supervision, strong cues to action and improved self-efficacy were all linked to increased PPE use, underscoring the importance of structured supervisory systems, continuous safety messaging and practical skill-building training. Conversely, perceived barriers such as discomfort, inconvenience and unavailability significantly reduced PPE uptake, indicating the need for ergonomic, climate-appropriate PPE and more efficient distribution mechanisms. The association between larger household size and PPE use suggests that family-centred messaging, emphasising how PPE protects income, prevents medical costs and preserves household stability, may enhance motivation and sustained adherence.

Broader systemic and cultural reforms are equally essential. Addressing stigma and elevating the professional status of sanitation workers requires a multimodal approach that integrates systems improvement, community education, workplace communication and cultural transformation. Collaborative platforms involving government agencies, OHS specialists, waste management innovators and community advocates can drive safer waste-handling practices, promote 3Rs (reduce, reuse, recycle) behaviours and support workforce dignity. Strengthening operational logistics, establishing supply chain tracking systems and formalising job descriptions will further professionalise the sector and reinforce the critical role sanitation workers play in public health and urban resilience. Ultimately, investing in the safety and dignity of sanitation workers is not only a policy obligation but a profound public health imperative that directly strengthens the resilience of communities and health systems.

Conclusions

This mixed-methods study provides comprehensive evidence that sanitation workers in Addis Ababa experience substantial gaps in access to safe sanitation services and PPE, exposing them to considerable OHS risks. PPE utilisation was associated with household size, workplace PPE availability, supervision, cues to action, self-efficacy and perceived barriers. The qualitative findings complemented these results by demonstrating that inconsistent PPE provision, inadequate occupational health training, weak enforcement of OHS regulations, precarious employment conditions and limited institutional support represent important organisational barriers to safe work practices. Together, these findings suggest that improving occupational safety among sanitation workers requires integrated interventions that address both organisational systems and behavioural determinants. Strengthening reliable PPE procurement and replacement systems, regular occupational health training, supportive supervision and effective enforcement of OHS regulations may improve access to safe sanitation services and promote consistent PPE utilisation. Prioritising sanitation workers within national occupational health and sanitation policies is essential to protect this underserved workforce, reduce occupational health risks and strengthen resilient sanitation and public health systems.

All authors are affiliated with the University of Gondar, which provided institutional support for this research. The University of Gondar had no role in the study design; data collection, analysis or interpretation; manuscript preparation or the decision to submit the manuscript for publication. The views expressed in this article are solely those of the authors and do not necessarily represent the official views of the University of Gondar.

Data availability statement

Data are available upon reasonable request.