Transformational Leadership, Work Engagement, and Patient Safety Culture in Public and Private Hospitals in Mogadishu, Somalia: A Study Protocol and Statistical Mediation Analysis

Abstract

Background

Effective hospital leadership can influence workforce engagement, teamwork, communication, learning from error, and patient safety, but evidence from fragile and resource-constrained health systems remains limited.

Objective

To present a study protocol for examining the association between transformational leadership and patient safety culture, and for testing healthcare workers’ engagement as a statistical mediating (indirect) pathway between the two, in public and private hospitals in Mogadishu, Somalia.

Methods

The protocol specifies a multicentre, institution-based analytical cross-sectional design with a target sample of 300 healthcare professionals selected through multistage stratified sampling, using a questionnaire that measures transformational leadership, work engagement, and ten dimensions of patient safety culture. To demonstrate the planned analytical workflow, this document also reports an illustrative analysis based on 300 computer-generated, synthetic responses; these figures are not empirical data collected from healthcare workers and are presented solely for methodological demonstration.

Illustrative Results

In the synthetic dataset, the mean transformational leadership score was 3.09 (SD 0.84), work engagement was 3.38 (SD 1.38), and patient safety culture was 3.19 (SD 0.81). Leadership was correlated with engagement (r=0.420) and patient safety culture (r=0.465), and the illustrative indirect effect of leadership on patient safety culture through engagement was 0.137 (bootstrap 95% CI 0.089 to 0.187).

Conclusion

The synthetic analysis demonstrates the proposed analytical workflow for interpreting leadership, engagement, and patient safety culture together. These figures are not empirical findings from Somali hospitals; substantive conclusions require implementation of the protocol with real participants.

Background

Patient safety is a fundamental dimension of healthcare quality and a prerequisite for effective, people-centred health services. Avoidable harm may arise from medication errors, diagnostic delays, communication failures, unsafe procedures, inadequate staffing, and weaknesses in organizational systems. The World Health Organization’s Global Patient Safety Action Plan 2021–2030 calls on health systems to eliminate avoidable harm through leadership commitment, organizational learning, workforce capability, patient engagement, and reliable safety processes.¹ Consequently, patient safety should be understood not only as the responsibility of individual clinicians but also as an outcome of leadership, governance, organizational culture, and the design of care systems.

Hospital leaders shape the conditions under which healthcare professionals deliver care. Their decisions influence staffing, workload, resource allocation, staff development, communication, incident reporting, and the response to adverse events. Leaders also determine whether employees experience the organization as supportive and fair or punitive and blame-oriented. When leadership encourages openness, teamwork, learning, and psychological safety, staff members may be more willing to identify risks and report near misses. Conversely, weak communication, inconsistent supervision, limited feedback, and fear of punishment can suppress reporting and allow preventable safety problems to recur.^2–6

Transformational leadership is especially relevant to healthcare because it extends beyond routine supervision and transactional exchange. It involves communicating a shared vision, acting as a credible professional example, recognizing staff contributions, supporting individual development, encouraging participation, and stimulating innovative problem-solving.⁷ These behaviours may help clinical teams align daily practice with broader goals for quality and safety. Systematic-review evidence indicates that transformational leadership is associated with more supportive nursing work environments, stronger staff performance, greater job satisfaction, improved quality of care, and more favourable patient-related outcomes.^2,3

Empirical studies also suggest that transformational leadership can influence patient safety through several organizational pathways. Boamah et al reported that nurse managers’ transformational leadership was associated with workplace empowerment, job satisfaction, and patient safety outcomes.⁴ Hamdan et al found positive relationships among transformational leadership, patient safety culture, and nurses’ safety practices, with patient safety culture operating as a mediator.⁵ Research in community and long-term-care services has similarly linked transformational leadership with work engagement, job resources, person-centred care, and stronger safety culture.⁶ Although these findings support the relevance of leadership, evidence from fragile and resource-constrained hospital systems remains limited.

According to job demands-resources theory, work engagement is fostered when employees perceive sufficient job resources, such as supportive leadership, autonomy, and recognition, relative to job demands, and engaged employees are in turn more likely to invest effort in safe, high-quality care.⁸ Healthcare workers’ engagement may therefore be an important mechanism connecting leadership to patient safety culture. Work engagement is a positive work-related state characterized by vigour, dedication, and absorption.⁹ Engaged staff are more likely to invest physical, cognitive, and emotional effort in their work, communicate proactively, support colleagues, and participate in quality-improvement activities. Transformational leaders may strengthen engagement by providing meaning, recognition, autonomy, professional development, and confidence in organizational goals. In contrast, persistent workload pressure, poor staffing, limited recognition, and exclusion from decision-making may weaken engagement and reduce employees’ willingness to speak up about safety concerns.^4–6

Patient safety culture refers to shared values, beliefs, norms, and behavioural expectations concerning the prevention of harm. It includes teamwork, communication openness, organizational learning, non-punitive response to error, leadership support, incident reporting, staffing, and effective information exchange. The AHRQ Hospital Survey on Patient Safety Culture Version 2.0 provides a structured approach for assessing these dimensions among hospital staff.^10,11 Examining specific safety-culture dimensions is valuable because an institution may perform well in teamwork while still experiencing weaknesses in staffing, handoffs, event reporting, or management support.

The relevance of these relationships is heightened in Somalia, where the health system continues to recover from prolonged conflict, institutional fragmentation, workforce shortages, limited public financing, and uneven regulatory capacity. Somalia’s Health Sector Strategic Plan III identifies governance, leadership, quality of care, health workforce development, information systems, and institutional accountability as priorities for progress toward universal health coverage and health security.¹² Strengthening hospital leadership is therefore not separate from health-system reconstruction; it is one of the mechanisms through which policies, standards, and resources can be translated into safer clinical practice.

Peer-reviewed empirical evidence on transformational leadership, healthcare workers’ engagement, and patient safety culture specifically from Somali hospitals could not be identified; this absence is itself part of the evidence gap that the present protocol seeks to address. Regional evidence from neighbouring East African health systems illustrates comparable patterns using the same AHRQ instrument: hospital-based surveys in Ethiopia have reported overall positive patient safety culture scores of approximately 46% to 51%, with teamwork typically the strongest composite and staffing or event reporting among the weakest, and one comparative study reported significantly lower patient safety culture scores in private than in government facilities.^13,14 These findings suggest that hospital ownership, staffing adequacy, and incident-reporting systems are relevant organizational correlates of patient safety culture in comparable fragile and transitional health systems, and they reinforce the rationale for examining these relationships empirically in Mogadishu.

Mogadishu contains a diverse mix of public hospitals, private for-profit facilities, private not-for-profit organizations, and externally supported services. The private sector delivers a substantial proportion of curative care, but facilities differ in management arrangements, staffing patterns, financing, data systems, and quality-assurance mechanisms.¹⁵ Public hospitals in Mogadishu typically operate under centralized government budgeting, civil-service staffing rules, and externally supported reporting structures, whereas private facilities generally have more autonomous management, fee-based financing, and variable internal quality-assurance systems; these structural differences may influence supervisory continuity, staff accountability, resource availability, and the extent to which incident reporting is protected from punitive consequences. Comparable research in Ethiopia found that patient safety culture scores differed significantly by hospital ownership.¹⁴ These differences may therefore affect how employees experience leadership, engagement, and patient safety. Comparing public and private hospitals is accordingly treated as an explicit objective of this protocol (Objective 4), providing evidence about shared challenges and context-specific organizational strengths.

Reliable organizational data are also essential for leadership and patient-safety improvement. A comprehensive assessment of Somalia’s health information system identified fragmented governance, variable standards, limited workforce capacity, infrastructure constraints, and insufficient use of data for decision-making.¹⁶ At hospital level, similar weaknesses may limit the systematic monitoring of incidents, staff feedback, quality indicators, and corrective actions. Measuring staff perceptions can complement routine administrative information by identifying organizational conditions that may not be captured through existing reporting 2Despite the policy importance of leadership and patient safety, multicentre evidence from Somali hospitals remains scarce. Previous work has focused more frequently on access, financing, service availability, workforce capacity, and health-information systems than on the relationships among leadership behaviour, employee engagement, and safety culture. It is therefore unclear whether healthcare workers who perceive stronger transformational leadership also report greater work engagement and a more positive patient safety culture, or whether these relationships differ between public and private institutions.

This study is designed to address that evidence gap by assessing transformational leadership, healthcare workers’ engagement, and patient safety culture in selected public and private hospitals in Mogadishu. It will estimate the adjusted association between leadership and patient safety culture and examine whether work engagement constitutes an indirect statistical pathway linking the two constructs. The findings may guide hospital executives, clinical managers, policymakers, professional councils, and development partners in designing leadership-development initiatives, workforce-engagement strategies, non-punitive reporting systems, and targeted quality-improvement programmes. Because the proposed design is cross-sectional, mediation findings will be interpreted as indirect statistical associations rather than proof of temporal or causal mechanisms.

Study Objectives

General Objective

To examine the association between transformational leadership and patient safety culture and assess the mediating role of healthcare workers’ engagement in public and private hospitals in Mogadishu, Somalia.

Specific Objectives

1. To assess the levels of transformational leadership, healthcare workers’ engagement, and patient safety culture in selected public and private hospitals in Mogadishu.

2. To determine the association between transformational leadership and patient safety culture after controlling for relevant individual and workplace characteristics.

3. To examine whether healthcare workers’ engagement statistically mediates the association between transformational leadership and patient safety culture.

4. To compare mean levels of transformational leadership, healthcare workers’ engagement, and patient safety culture between public and private hospitals in Mogadishu.

Methodology

Study Design

A multicentre, institution-based analytical cross-sectional study will be conducted. The design will estimate the distribution of the study constructs and test prespecified associations among transformational leadership, work engagement, and patient safety culture. Because the exposure, mediator, and outcome will be measured at the same time, any indirect effect will be interpreted as statistical mediation rather than proof of temporal or causal mediation.

Study Setting

The study will be undertaken in selected public and private hospitals in Mogadishu, Somalia. Hospitals will be eligible when they provide inpatient and outpatient services, have operated continuously for at least two years, employ at least 30 healthcare professionals, possess an identifiable clinical and administrative leadership structure, and grant institutional permission. Stand-alone clinics, diagnostic centres, temporary facilities, and hospitals without an accessible staff roster will be excluded.

Study Population

The source population will consist of healthcare professionals employed in eligible hospitals in Mogadishu. The study population will comprise randomly selected physicians, nurses, midwives, pharmacists, laboratory professionals, anaesthesia professionals, radiology personnel, nutrition professionals, physiotherapists, and other allied healthcare workers who meet the eligibility criteria.

Inclusion Criteria

Participants must be at least 18 years old, have worked in the current hospital for at least six months, have an identifiable immediate supervisor or clinical leader, and provide informed consent. Six months of employment is considered sufficient minimum exposure to the hospital’s leadership and safety environment.

Exclusion Criteria

Students, interns, unpaid trainees, temporary workers with less than six months of service, staff on extended leave, chief executives or hospital owners without an immediate supervisor, and individuals who participated in the pilot study will be excluded.

Sample Size Determination

The sample size is calculated using Cochran’s single-population proportion formula with a 95% confidence level, an assumed proportion of 50%, and a 6% margin of error because no reliable prior estimate of positive patient safety culture is available for the target population.

This calculation addresses estimation precision for the primary descriptive objective (the overall proportion of positive patient safety culture) rather than statistical power for the adjusted regression and mediation analyses. As an approximate check for the planned analyses, a target of 300 respondents provides at least 20 observations per covariate for the adjusted regression model (nine covariates) and exceeds commonly used minimum samples (roughly n=200) for bootstrap-based single-mediator models with medium effect sizes; this is reported as a pragmatic adequacy check rather than a formal power calculation for mediation.

n₀ = Z²p(1 − p) / d²

Using Z=1.96, p=0.50, and d=0.06 gives an initial sample of approximately 267 participants. After adding approximately 10% for non-response or incomplete questionnaires, the final target is rounded to 300 healthcare professionals. The parameters used in this calculation are summarized in Table 1.

Table 1.

Parameters Used in the Sample Size Calculation

Parameter	Value
Confidence level	95% (Z=1.96)
Assumed proportion	50%
Margin of error	6%
Initial sample	267
Final target sample	300

Sampling Procedure

A multistage stratified sampling procedure will be used. Eligible hospitals will first be stratified into public and private institutions, and each stratum will be listed and numbered. Four public and four private hospitals will be selected from these lists by computer-generated simple random sampling; if a selected hospital declines to participate or does not grant institutional permission, the next hospital on the same randomly ordered list will be approached as a replacement, and the reason for any substitution will be documented. The total sample will be allocated proportionally to the number of eligible healthcare professionals in each selected hospital, using the allocation formula below, and participants will be selected from updated staff lists by computer-generated simple random sampling. The covariates listed in Study Variables (age, sex, profession, experience, working hours, training, staffing adequacy, and hospital ownership) will be collected for all participants and used to describe the sample and to adjust the regression and mediation models; no additional matching or quota criteria will be applied at the hospital-selection stage.

n_h = (N_h / N) × 300

Study Variables

The study variables, their roles, and their measurement are summarized in Table 2.

Table 2.

Study Variables and Their Measurement

Variable role	Variable	Measurement
Independent variable	Transformational leadership	Seven items scored from 1 to 5.
Mediating variable	Healthcare workers’ engagement	Nine items scored from 0 to 6.
Primary outcome	Patient safety culture	Thirty-two items forming ten composite measures.
Covariates	Individual and workplace factors	Age, sex, profession, experience, working hours, training, staffing, hospital ownership, and related characteristics.

Data Collection Instrument

Data will be collected using an anonymous, structured, self-administered questionnaire containing four components. The instrument combines investigator-developed background questions with established measures of transformational leadership, work engagement, and patient safety culture:

• Sociodemographic and occupational characteristics.

• Transformational leadership behaviours adapted from the seven-item Global Transformational Leadership scale developed by Carless, Wearing, and Mann.¹⁷

• Work engagement adapted from the nine-item Utrecht Work Engagement Scale (UWES-9).⁹

• Patient safety culture based on the AHRQ Surveys on Patient Safety Culture Hospital Survey Version 2.0 and its official scoring guidance.^10,11

The research team must verify all current instrument-use conditions, permissions, translation requirements, and attribution requirements before field administration. When the official AHRQ SOPS Hospital Survey 2.0 is administered as an AHRQ instrument, the core item wording, response options, definitions, and order should follow the official materials.^10,11

Translation and Cultural Adaptation

The English questionnaire will be independently translated into Somali by two bilingual translators with knowledge of healthcare terminology. A third translator, blinded to the original, will back-translate the reconciled Somali version. An expert committee will assess semantic, conceptual, and cultural equivalence, followed by cognitive interviews with 10–15 healthcare workers in accordance with established cross-cultural adaptation principles.¹⁸

Pilot Study

The questionnaire will be piloted among approximately 30 healthcare professionals in a hospital not included in the main study. The pilot will assess clarity, completion time, acceptability, missing responses, administration procedures, and preliminary reliability. Pilot participants will not be included in the main analysis.

Validity and Reliability

Content validity will be assessed by an expert panel prior to piloting. Internal consistency will be examined using Cronbach’s alpha and McDonald’s omega, with coefficients of at least 0.70 generally considered acceptable Confirmatory factor analysis of the translated instruments, conducted in R (lavaan package) or an equivalent structural-equation-modelling program, will be performed once main-study data collection is complete and before hypothesis testing, to evaluate the structure of the scales in the Somali hospital context; results will be reported alongside the substantive findings rather than used to select items post hoc.

Data Collection Procedure

Trained research assistants who are not part of participating hospitals’ management teams will administer the survey. Hospital administrators will provide staff lists and logistical access but will not observe questionnaire completion or view individual responses. Participants will complete the questionnaire privately and return it in a sealed envelope or through a secure electronic form.

Data Quality Assurance

• Standardized training and written instructions for data collectors.

• Pilot testing and correction of unclear wording.

• Daily review for completeness without recording personal identifiers.

• Double data entry or programmed range and consistency checks.

• A documented data-cleaning log and secure, access-controlled storage.

• Anonymous reporting to reduce social-desirability and fear-of-retaliation bias.

Operational Definitions and Scoring

The operational definitions and scoring rules for the main study constructs are presented in Table 3.

Table 3.

Operational Definitions and Scoring of the Study Constructs

Construct	Operational Definition
Transformational leadership	Mean of B1-B7; range 1–5. Higher scores indicate stronger perceived leadership.
Work engagement	Mean of C1-C9; range 0–6. Higher scores indicate greater energy, dedication, and absorption.
Patient safety culture	Negatively worded items are reverse-coded. Ten composite means and percent-positive scores are calculated.
Positive patient safety response	A response of 4 or 5 on positively worded items, or 1 or 2 on negatively worded items.

Statistical Analysis

Data will be analysed using IBM SPSS Statistics (descriptive statistics, reliability analysis, correlation, and multiple linear regression) and R, version 4.x (packages boot and lavaan, for bootstrap mediation and confirmatory factor analysis); the exact software versions will be fixed before data collection begins and reported in the final study report. Categorical variables will be summarized using frequencies and percentages; continuous variables will be summarized using means and standard deviations or medians and interquartile ranges.

The primary outcome is the overall patient safety culture score, calculated as the unweighted mean of the ten composite dimension scores described in Operational Definitions and Scoring (equivalently, the mean of all reverse-coded items). This overall score is the outcome used in the adjusted regression and mediation models. The ten dimension-specific composite scores (reported in Patient Safety Culture Composite Results) are analysed descriptively, by construction, and are treated as secondary, exploratory outcomes rather than separate confirmatory endpoints.

Scale reliability will be assessed before composite scores are formed. Pearson or Spearman correlations will examine unadjusted relationships among leadership, engagement, and patient safety culture. Public and private hospitals will be compared using appropriate independent-samples tests and adjusted regression models.

Multiple linear regression will estimate the adjusted association between transformational leadership and patient safety culture. Covariates will be selected using the conceptual framework rather than only bivariate statistical significance.

Mediation will be examined using a regression-based model equivalent to PROCESS Model 4. The indirect effect will be calculated as path a multiplied by path b and estimated with 5000 bootstrap resamples. An indirect association will be considered statistically supported when the 95% bootstrap confidence interval does not include zero.

Findings will be reported with coefficients, effect sizes, 95% confidence intervals, exact p-values, and model diagnostics. Interpretation will emphasize practical relevance and the limitations of cross-sectional mediation.

Management of Missing Data

The amount and pattern of missing data will be described for all study variables. Scale scores will be calculated only when at least 80% of the constituent items are completed. Multiple imputation by chained equations (R package mice, or the equivalent SPSS procedure) will be performed if missingness on any covariate exceeds 5% and is plausibly missing at random; imputation, if performed, will be carried out after data cleaning and before the regression and mediation analyses, and complete-case results will be reported alongside the imputed results as a sensitivity analysis.

Ethical Considerations

Ethical approval will be obtained from the responsible Institutional Review Board or Research Ethics Committee before data collection. Administrative permission will be obtained from every participating hospital. Written or approved electronic informed consent will be obtained. No names, employee numbers, or direct identifiers will be recorded, and only aggregate results will be shared.

Reporting

The completed empirical study will be reported according to STROBE recommendations for cross-sectional studies.¹⁹ Prespecified and exploratory analyses will be distinguished, and the limitations of self-report and cross-sectional mediation will be stated clearly.

Illustrative Results from 300 Synthetic Responses

All numerical results in Illustrative Results From 300 Synthetic Responses are generated from fictional records. They demonstrate how real results may be organized and interpreted, but they are not empirical findings.

Participant Characteristics

The synthetic dataset contained 300 records from eight fictional hospitals, with equal allocation between public and private hospitals. The mean age was 35.0 years (SD 7.4); 175 (58.3%) respondents were female and 125 (41.7%) were male. Selected characteristics of the synthetic respondents are presented in Table 4.

Table 4.

Selected Characteristics of the 300 Synthetic Respondents

Variable	Category	n	%
Sex	Female	175	58.3
	Male	125	41.7
Hospital ownership	Public	150	50.0
	Private for-profit	150	50.0
Safety training in past 24 months	Yes	175	58.3
	No	125	41.7
Leadership training in past 24 months	Yes	92	30.7
	No	208	69.3
Staffing adequacy	Very inadequate	54	18.0
	Inadequate	69	23.0
	Adequate	114	38.0
	Very adequate	63	21.0
Shift pattern	Day only	85	28.3
	Night only	21	7.0
	Rotating shifts	189	63.0
	Other	5	1.7

Notes: Author’s own computation, based on the illustrative synthetic dataset generated for this manuscript; values are simulated for demonstration purposes and do not represent data collected from actual respondents.

Main Study Constructs and Reliability

The scales showed strong internal consistency in the synthetic dataset (Table 5). Transformational leadership had a mean of 3.09, work engagement had a mean of 3.38, and patient safety culture had a mean of 3.19 (Figure 1).

Table 5.

Descriptive Statistics and Internal Consistency

Construct	Mean	SD	Minimum	Maximum	Cronbach α
Transformational leadership	3.09	0.84	1.00	5.00	0.837
Work engagement	3.38	1.38	0.00	6.00	0.905
Patient safety culture	3.19	0.81	1.28	4.87	0.963

Notes: Author’s own computation, based on the illustrative synthetic dataset generated for this manuscript; values are simulated for demonstration purposes and do not represent data collected from actual respondents.

Figure 1.

Mean scores of the main study constructs (synthetic data).

Notes: Author’s own construction, based on the illustrative synthetic dataset generated for this manuscript; values are simulated for demonstration purposes and do not represent data collected from actual respondents.

Correlations Among the Main Constructs

As shown in Table 6, transformational leadership was positively correlated with work engagement (r=0.420) and patient safety culture (r=0.465). Work engagement was also positively correlated with patient safety culture (r=0.529).

Table 6.

Correlation Matrix

Construct	1	2	3
1. Transformational leadership	1.000	0.420	0.465
2. Work engagement	0.420	1.000	0.529
3. Patient safety culture	0.465	0.529	1.000

Notes: Author’s own computation, based on the illustrative synthetic dataset generated for this manuscript; values are simulated for demonstration purposes and do not represent data collected from actual respondents.

Comparison by Hospital Ownership

Mean scores of the three main constructs in public and private hospitals are compared in Table 7 and Figure 2. In the synthetic dataset, private hospitals had higher mean scores than public hospitals for transformational leadership (3.20 vs 2.98; p=0.028) and patient safety culture (3.31 vs 3.06; p=0.007), whereas the difference in work engagement was not statistically significant (3.45 vs 3.31; p=0.368).

Table 7.

Public-Private Comparison of Mean Scores

Construct	Public n	Public Mean	Public SD	Private n	Private Mean	Private SD	t	p
Transformational leadership	150	2.98	0.86	150	3.20	0.81	−2.210	0.0279
Work engagement	150	3.31	1.34	150	3.45	1.42	−0.902	0.3676
Patient safety culture	150	3.06	0.82	150	3.31	0.79	−2.697	0.0074

Notes: Author’s own computation, based on the illustrative synthetic dataset generated for this manuscript; values are simulated for demonstration purposes and do not represent data collected from actual respondents.

Figure 2.

Mean construct scores by hospital ownership (synthetic data).

Notes: Author’s own construction, based on the illustrative synthetic dataset generated for this manuscript; values are simulated for demonstration purposes and do not represent data collected from actual respondents.

Regression and Statistical Mediation

The illustrative adjusted regression and mediation results are presented in Table 8. The synthetic indirect effect was 0.137, with a bootstrap 95% confidence interval from 0.089 to 0.187. Because this interval did not include zero, the simulated results demonstrate a statistically supported indirect association through work engagement. This does not establish causal mediation.

Table 8.

Illustrative Adjusted Regression and Mediation Analysis

Path/Effect	B	SE	95% CI Lower	95% CI Upper	p	R2
Path a: Leadership → Engagement	0.692	0.085	0.526	0.858	0.0000	0.254
Total effect c: Leadership → Safety culture	0.443	0.048	0.349	0.536	0.0000	0.322
Direct effect c′: Leadership → Safety culture	0.305	0.049	0.208	0.402	0.0000	0.407
Path b: Engagement → Safety culture	0.198	0.031	0.138	0.259	0.0000	0.407
Indirect effect a×b	0.137	—	0.089	0.187	—	—

Notes: Author’s own computation, based on the illustrative synthetic dataset generated for this manuscript; values are simulated for demonstration purposes and do not represent data collected from actual respondents.

Patient Safety Culture Composite Results

The patient safety culture composite results are presented in Table 9 and Figure 3. In the synthetic dataset, overall percent-positive scores ranged from 39.3% for Event Reporting to 46.0% for Supervisor Support.

Table 9.

Patient Safety Culture Composite Results

Composite	Mean	Overall Positive %	Public %	Private %	Rank (1=highest)
Teamwork	3.14	40.1	34.2	46.0	9 of 10
Staffing and Work Pace	3.21	42.2	39.0	45.3	6 of 10
Organizational Learning	3.22	42.9	36.9	48.9	4 of 10
Response to Error	3.18	42.4	39.8	45.0	5 of 10
Supervisor Support	3.22	46.0	39.6	52.4	1 of 10
Communication About Error	3.21	43.1	38.9	47.3	3 of 10
Communication Openness	3.15	41.7	37.2	46.2	8 of 10
Event Reporting	3.12	39.3	36.3	42.3	10 of 10
Management Support	3.22	43.3	38.2	48.4	2 of 10
Handoffs and Information Exchange	3.20	42.1	39.8	44.4	7 of 10

Notes: Author’s own computation, based on the illustrative synthetic dataset generated for this manuscript; values are simulated for demonstration purposes and do not represent data collected from actual respondents. The rank column orders composites from highest (1) to lowest (10) by overall percent-positive score within this illustrative dataset only; it is a descriptive ordering, not a validated benchmark or a classification of priority.

Figure 3.

Patient safety culture percent-positive results (synthetic data).

Notes: Author’s own construction, based on the illustrative synthetic dataset generated for this manuscript; values are simulated for demonstration purposes and do not represent data collected from actual respondents.

Interpretation and Discussion

The illustrative dataset produced a mean transformational leadership score of 3.09 (SD 0.84, range 1–5), a mean patient safety culture score of 3.19 (SD 0.81, range 1–5), and a mean work engagement score of 3.38 (SD 1.38, range 0–6). The reliability coefficients were above 0.80 for all principal scales, suggesting that the item sets functioned consistently within the generated data. In a real study, these estimates would need to be interpreted alongside factor analysis and item-level diagnostics.

Transformational leadership was positively related to both work engagement and patient safety culture. This pattern is consistent with leadership theory, which proposes that leaders may influence quality and safety by communicating a shared vision, supporting staff development, encouraging participation, and establishing psychologically safe conditions for speaking up and learning from error.^2–5

Work engagement was positively associated with patient safety culture after leadership and selected covariates (age, sex, profession, experience, working hours, training, staffing adequacy, and hospital ownership) were controlled. The indirect effect through engagement was 0.137. This simulated pattern illustrates a possible workforce pathway: supportive leadership may strengthen employees’ energy and dedication, which may in turn facilitate teamwork, attention to risk, communication, and participation in quality improvement.

The highest-ranked synthetic patient safety composite was Supervisor Support (46.0% positive; rank 1 of 10), whereas Event Reporting was the lowest-ranked (39.3% positive; rank 10 of 10). In an empirical study, low scores in staffing, work pace, handoffs, communication openness, or event reporting would require targeted organizational responses rather than individual blame.

Differences between public and private hospitals should be interpreted cautiously. This comparison corresponds to the prespecified objective of comparing public and private hospitals (Objective 4) rather than an exploratory analysis. Hospital ownership may represent differences in financing, staffing, management systems, patient volume, reporting infrastructure, and workforce composition. A larger number of hospitals and multilevel modelling would be preferable for separating individual-level and hospital-level influences.

The use of a cross-sectional design means that the temporal order of leadership, engagement, and safety culture cannot be established. It is also possible that employees working in stronger safety environments perceive their leaders and their own engagement more positively. Longitudinal, mixed-methods, and intervention studies would provide stronger evidence about mechanisms and change over time.

Conclusion

This integrated document presents a complete protocol, questionnaire, coding framework, synthetic response matrix, and illustrative analytical workflow for examining transformational leadership, work engagement, and patient safety culture in Mogadishu hospitals. The simulated analysis demonstrates how direct and indirect associations can be reported. Actual conclusions about Somali hospitals can be made only after ethical approval, real participant recruitment, data collection, validation, and analysis. Recommendations: 1. Obtain formal ethical approval and written permission from every participating hospital before recruitment. 2. Confirm permission and approved wording for all measurement instruments and complete Somali translation and cultural adaptation. 3. Use probability sampling and document the number of eligible staff, selected staff, respondents, and non-respondents in every hospital. 4. Protect anonymity by ensuring that managers cannot view individual questionnaires or identify respondents from small departments. 5. Conduct reliability and factor analyses before interpreting aggregate scale scores. 6. Use the synthetic workbook and response matrix only for training; replace every simulated value with verified empirical data before preparing a submission. 7. Report cross-sectional mediation as an indirect statistical association and not as proof of causality.

Strengths and Limitations

Strengths

• The protocol addresses a documented evidence gap: multicentre data linking transformational leadership, work engagement, and patient safety culture are scarce for fragile, resource-constrained health systems such as Somalia’s.

• The design compares public and private hospitals within the same city, which helps separate ownership-related effects from broader contextual factors.

• All three constructs are measured with instruments that have established validity and reliability in other settings (the Global Transformational Leadership scale, UWES-9, and AHRQ SOPS Hospital Survey 2.0), rather than investigator-improvised items.

• The mediation pathway through work engagement is prespecified rather than explored post hoc, and the plan explicitly commits to interpreting it as an indirect statistical association rather than a causal claim, given the cross-sectional design.

• The inclusion of a full questionnaire, coding guide, and dry-run analysis workflow means the study can move directly from ethical approval to data collection and analysis without further instrument development.

Limitations

• The present numerical results are based entirely on synthetic, computer-generated responses and have no empirical validity.

• The proposed cross-sectional design cannot establish temporal sequence or causal mediation.

• Self-reported leadership, engagement, and safety culture may be affected by social-desirability and common-method bias.

• Healthcare workers are clustered within hospitals; analysis should account for clustering and the limited number of institutions.

• Translated instruments require formal cultural adaptation and psychometric evaluation in the Somali context.

• Findings from hospitals in Mogadishu may not be generalizable to other regions of Somalia.

Funding Statement

This research received fund from SIMAD University.

Data Sharing Statement

All data supporting the findings of this study are available from the corresponding author upon reasonable request.

Ethics Approval and Consent to Participate

The application for ethics clearance for the above-mentioned research was reviewed by ethical approval committee SIMAD University on (09/04//2026) in compliance with the SIMAD University policy on Research Ethics. The Study period was 3 Months. Ethical clearance has been granted. Please be advised that the research ethics committee needs to be informed should any part of the research methodology as outlined in the ethics Application (Ref: 2026/SU-IRB/P00293), change in any way. The Research Ethics Review Committee wishes you all the best with this research undertaking.

Author Contributions

All authors made a significant contribution to the work reported, whether that is in the conception, study design, execution, acquisition of data, analysis and interpretation, or in all these areas; took part in drafting, revising or critically reviewing the article; gave final approval of the version to be published; have agreed on the journal to which the article has been submitted; and agree to be accountable for all aspects of the work.

Disclosure

The authors declare that there are no financial, personal, or professional conflicts of interest that could have influenced the work reported in this study. No funding or sponsorship was received that could have affected the study design, data collection, analysis, interpretation, or publication of the results.