Research Engagement and Evidence Utilization Among Undergraduate-Trained Nurses in Kenyan National Referral Hospitals: A Mixed-Methods Investigation

Introduction

Research involvement and evidence utilization represent two interconnected pillars of professional nursing competence in contemporary healthcare systems. Together, they constitute the operational foundation of evidence-based practice (EBP)defined as the deliberate integration of the best available research evidence with clinical expertise and patient values to optimize health outcomes (Melnyk & Fineout-Overholt, 2019; Sackett et al., 1996). Research involvement encompasses a broad spectrum of activities extending from original knowledge generation including primary data collection, analysis, and dissemination to the active utilization of research findings produced by others in the context of clinical decision-making (Estabrooks, 1999; Squires et al., 2011). Both dimensions are indispensable: a nursing profession that only produces research without applying it, or that only consumes research without contributing to its generation, cannot sustain the dynamic knowledge cycle that EBP demands. In Kenya, the significant expansion of Bachelor of Science in Nursing (BScN) degree programs since the early 2000s has been explicitly framed within national policy as a strategy for cultivating a graduate nursing workforce embedded in research culture (Ministry of Health Kenya, 2012; Nursing Council of Kenya [NCK], 2019). The Kenya Health Policy 2012-2030 identifies strengthening of health research and its application in clinical practice as a core pillar of health system transformation (Ministry of Health Kenya, 2012). Within this policy framework, BScN graduate shaving received formal undergraduate instruction in research methodology, biostatistics, critical appraisal, and EBP principles are expected to serve as research-engaged change agents within their clinical institutions, actively contributing to both the generation and utilization of nursing knowledge. However, empirical evidence from sub-Saharan Africa presents a considerably more complex picture. Multiple studies across diverse African healthcare contexts have documented persistently low rates of post-graduation research activity, passive rather than active evidence consumption, and a substantial chasm between espoused EBP values and enacted EBP behaviors (Abubakar et al., 2020; Olade, 2004; Perez et al., 2021). A constellation of structural barriers encompassing inadequate library and database access, unreliable internet infrastructure, high patient-to-nurse ratios, chronic understaffing, absent mentorship structures, and hierarchical interprofessional cultures has been consistently identified as a determinant of low research engagement among African nurses (Forsman et al., 2012; Saunders et al., 2019). These structural barriers interact with individual-level factors including perceived research complexity, limited statistical competence, and professional role identity to create a complex, multilayered impediment to sustained research involvement and robust evidence utilization. In Kenya's national referral hospital system characterized by high patient complexity, diverse pathological presentations, multi-specialty clinical teams, and formal training mandates the deficit in research involvement and the quality of evidence utilization among BScN-trained nurses carries direct implications for patient safety, clinical quality, health system efficiency, and the professional stature of nursing as a discipline (Njiru, 2018; Nzengya et al., 2023). Despite this significance, rigorous, multi-institutional empirical characterization of research involvement and evidence utilization patterns among Kenyan BScN nurses in tertiary settings remains limited. The extant literature is largely dominated by small convenience samples, non-validated instruments, and single-institution designs that restrict generalizability and the capacity for policy-relevant inference. This study addresses this evidence gap through a comprehensive mixed-methods investigation of research involvement and evidence utilization among BScN-trained nurses at two nationally designated teaching and referral hospitals: Moi Teaching and Referral Hospital (MTRH) in Eldoret and Nakuru County Teaching and Referral Hospital (NCTRH) in Nakuru.

Theoretical Framework

This study draws on two established theoretical frameworks to interpret patterns of research involvement and evidence utilization: the Diffusion of Innovations (DOI) Theory (Rogers, 2003) and the Promoting Action on Research Implementation in Health Services (PARIHS) framework (Kitson et al., 1998; Rycroft-Malone et al., 2002). The DOI theory (Rogers, 2003) conceptualizes the adoption of any innovation including evidence-based clinical practices as a process determined by five perceived attributes of the innovation itself and by the social system within which diffusion occurs. These five attributes are: relative advantage (whether EBP is seen as superior to current practice), compatibility (whether EBP fits existing values and workflows), complexity (how difficult EBP is to understand and use), trialability (whether nurses can experiment with EBP on a limited basis), and observability (whether the results of EBP use are visible to others). When EBP is perceived as too complex, insufficiently compatible with demanding clinical workflows, or offering limited observable benefit relative to the effort required, adoption rates decline and simpler, more accessible substitutes (such as SOPs or Google searches) are rationalized as pragmatic alternatives. This DOI prediction maps directly onto the adaptive pragmatism documented qualitatively in this study. The PARIHS framework (Kitson et al., 1998; Rycroft-Malone et al., 2002) provides a complementary institutional-level lens, positing that successful evidence implementation is a function of the dynamic interaction among three core elements: evidence (the nature, quality, and credibility of the research), context (the organizational culture, leadership quality, and evaluation capacity of the healthcare setting), and facilitation (the availability of skilled agents who actively support evidence implementation). Applied to the Kenyan referral hospital setting, the PARIHS framework identifies critical deficits in all three elements: evidence may be available but inaccessible due to infrastructure constraints; context is characterized by hierarchical power structures and cultures that do not consistently support nurse-led inquiry; and facilitation mechanisms journal clubs, EBP committees, research mentors are largely absent from most clinical departments. Together, DOI and PARIHS provide a theoretically integrated account of why research involvement remains low and evidence utilization remains passive in this setting despite formal undergraduate education that is designed to produce research-engaged nursing professionals. We assessed the level of undergraduate-trained nurses' involvement in conducting and utilizing research in their clinical practice at two national referral hospitals in Kenya.

Methods

Study Design and Philosophical Orientation

This study employed a concurrent embedded mixed-methods design (Creswell & Plano Clark, 2018), characterized by the simultaneous collection and analysis of quantitative and qualitative data within a single data collection phase, with explicit assignment of primary status to one strand. The quantitative strand was designated the primary component (75% weighting) and employed a descriptive cross-sectional survey design to measure the frequency, scope, and patterns of research involvement and evidence utilization across the eligible nurse population. The qualitative strand (25% weighting) was embedded within the quantitative phase and adopted a phenomenological orientation (Polit & Beck, 2021) to explore the subjective meanings, lived experiences, and contextual determinants that explain the quantitative patterns observed. The research was grounded in a pragmatic worldview (Morgan, 2014), which privileges the research problem over allegiance to a singular philosophical paradigm and validates the combination of methods based on their fitness for purpose. A purely quantitative approach could identify what the patterns of research involvement and evidence utilization are but would fail to explain why those patterns persist especially the mechanisms through which structural barriers, cultural norms, and resource constraints translate into individual-level behaviors and adaptations. Conversely, a purely qualitative approach would generate rich, contextually dense narratives but could not establish population-level prevalence or test the distribution of behaviors and attitudes across the target population. The mixed-methods design transcends this limitation by capitalizing on the complementary strengths of both paradigms. Integration occurred during the interpretive phase, where quantitative patterns and qualitative themes were brought into systematic dialogue through a narrative weaving approach (O'Cathain et al., 2010). Qualitative themes were used to explain, elaborate, and contextualize quantitative findings, yielding a richer and more actionable meta-inference than either strand could produce independently.

Study Setting

The study was conducted at two purposively selected national teaching and referral hospitals that together represent distinct geographic and institutional contexts within Kenya's tertiary health system. Moi Teaching and Referral Hospital (MTRH), located in Eldoret, Uasin Gishu County, is a major national referral and teaching facility serving the western region of Kenya and operating in collaboration with several medical and nursing schools. Nakuru County Teaching and Referral Hospital (NCTRH), situated in Nakuru, functions as a regional referral hub for the Rift Valley region. Both institutions carry dual mandates for specialized clinical service delivery and health professions education, and both employ substantial numbers of BScN graduates engaged in direct clinical care, nursing administration, and nurse education roles (Ministry of Health Kenya, 2012; Njiru, 2018). Their dual service-and-teaching identity makes them ideal sites for investigating how the post-graduation professional environment shapes research engagement among graduate nurses.

Population and Sampling

The study targeted BScN-trained registered nurses formally employed at either MTRH or NCTRH on permanent, contractual, or long-term engagement terms with a minimum of one year of post-qualification clinical experience. This experience threshold was established to ensure participants had adequate exposure to routine clinical decision-making, quality improvement activities, and the institutional environments in which EBP is (or is not) operationalized (Polit & Beck, 2021). Applying these criteria to the combined workforce of 844 nurses across both facilities yielded an eligible population of 238 nurses. For the quantitative component, the National Education Association (1960) formula was applied at a 95% confidence level (χ² = 3.841, P = 0.5, d = 0.05), yielding a required sample of 147 nurses. Stratified random sampling was employed across clinical departments and professional role strata to ensure balanced representation. A total of 136 completed questionnaires were returned, yielding a response rate of 92.5%. For the qualitative component, 20 participants were purposively recruited 10 from each institution from eligible nurses not included in the quantitative survey. Participants represented clinical nurses, nurse administrators, and nurse educators, selected based on their potential to provide information-rich perspectives on research involvement and evidence utilization. Qualitative recruitment continued until thematic data saturation was achieved (Braun & Clarke, 2006; Guest et al., 2020).

Instrumentation

Questionnaire (RUQ; Champion & Leach, 1989; Estabrooks, 1999; Yoder et al., 2014). For this study, the RUQ was substantially adapted to assess three interconnected dimensions of research involvement: (a) research conduct including post-graduation research participation, purposes of research engagement, conference presentations, and peer-reviewed publications; (b) evidence utilization in practice including frequency, clinical contexts, evidence sources, and access tools; and (c) attitudinal dispositions toward research utilization, measured on a five-point Likert scale (1 = strongly disagree to 5 = strongly agree). The instrument also included a comprehensive demographic section. Reliability was confirmed through a pilot study with 15 BScN nurses at Langa Langa Hospital, Nakuru County, yielding Cronbach's α = .84-.92 across subscales, indicating good-to-excellent internal consistency. Qualitative data were collected using an unstructured interview guide comprising open-ended questions derived directly from the study objective. Interviews lasted 40-60 minutes and were conducted face-to-face or via video-enabled Zoom with participant consent. All sessions were audio-recorded and transcribed verbatim. Field notes captured non-verbal cues and contextual observations throughout.

Data Analysis

Quantitative data were entered and analyzed using IBM SPSS Version 25.0 at a significance threshold of α = .05. Descriptive statistics (frequencies and percentages) characterized research conduct patterns, evidence utilization behavior, and attitudinal distributions. Cross-tabulations examined the co-distribution of research involvement variables across demographic subgroups. Chi-square tests of independence (with Cramér's V effect size estimates) assessed associations between attitudinal items and subgroup variables (Kothari, 2004). Qualitative data were analyzed using the six-phase thematic analysis framework of Braun and Clarke (2006), proceeding from data familiarization through initial coding, theme development, theme review, theme definition, and report writing. Trustworthiness was ensured through Lincoln and Guba's (1985) criteria credibility, dependability, confirmability, and transferability operationalized through member checking, peer debriefing, audit trail maintenance, and reflexive journaling.

Ethical Considerations

Ethical approval was granted by the Moi Teaching and Referral Hospital Institutional Research and Ethics Committee (Ref No: IREC/2019/000090; Approval No: 0003491) and the National Commission for Science, Technology and Innovation (Ref No: NACOSTI/P/20/3063), with additional institutional approvals secured from both study hospitals. All participants provided written informed consent prior to data collection. Participants were assured of voluntary participation, the unconditional right to withdraw without penalty, and full anonymization through coded identification and secure data storage. Electronic data were password-protected; physical documents were maintained in locked filing cabinets. Audio recordings were permanently deleted following transcription verification.

Discussion

This study documents a striking and consequential paradox within the nursing workforce of Kenya's national referral hospitals: a pronounced active-passive dichotomy in which high self-reported evidence utilization (86.8%) coexists with low post-graduation research involvement (39.0%), and in which the predominant mode of evidence use is reactive compliance with institutional SOPs rather than proactive, critically appraised EBP. This finding challenges simplistic interpretations of EBP engagement as a binary present-or-absent phenomenon and instead reveals a nuanced landscape in which formal evidence use occurs within structural and cultural constraints that systematically limit its quality, depth, and professional sustainability. The finding that 61.0% of BScN nurses had not conducted any research since graduation despite universal undergraduate research educationist striking and yet theoretically predictable within both the DOI and PARIHS frameworks. Low post-graduation research conduct is not idiosyncratic to Kenya but is documented across comparable African nursing contexts, with rates ranging from 30% to 50% in the existing literature (Abubakar et al., 2020; Olade, 2004). The prominence of time constraints (47.0%) and resource deficits (28.9%) as leading barriers rather than motivational factors such as lack of interest (6.0%)is of critical importance for intervention design. It establishes that the barriers to research engagement are primarily structural rather than attitudinal, meaning that interventions targeting individual motivation or attitude change alone are unlikely to be effective without simultaneous addressing of the institutional and systemic conditions that create and perpetuate these structural barriers. The finding that academically motivated research (43.4%) substantially outpaced practice-change motivated research (30.2%) among those who did engage reveals that research activity among Kenya's graduate nurses is currently more strongly coupled to formal educational enrollment than to intrinsic professional identity. This pattern has important implications for sustainability: when the extrinsic academic motivational dissertation requirement, a postgraduate course, a supervisory expectation is removed, research activity ceases. Building an intrinsically motivated, professionally embedded research culture requires that research engagement be woven into the ongoing fabric of clinical work through protected time, institutional recognition, and career-linked incentives none of which are currently systematically available in the study hospitals (Forsman et al., 2012; Squires et al., 2011). While the high evidence utilization rate (86.8%) is superficially encouraging, the nature of that utilization heavily reliant on SOPs (76.3%) and Google (77.1%), with only 17.0% accessing indexed databases signals a fundamentally passive, compliance-oriented mode of evidence engagement rather than the active, critically appraised EBP that international guidelines and nursing education frameworks aspire to produce (Melnyk & Fineout-Overholt, 2019; WHO, 2020). This SOP-Google paradigm, while pragmatically rational in high-workload, infrastructure-constrained environments, represents a systematic downgrade in the rigor and currency of evidence guiding clinical decisions with direct consequences for care quality and patient safety in complex referral hospital environments. Estabrooks's (1999) taxonomy of research utilization illuminates this distinction: the evidence use documented quantitatively in this study most closely approximates symbolic utilization the use of evidence (or its institutional surrogates) to justify and legitimize pre-existing clinical decisions rather than instrumental utilization, which would entail the direct, specific application of research findings to clinical decision-making. The PARIHS framework explains this transition from the ideal (instrumental EBP) to the actual (symbolic compliance) as a predictable consequence of weak context and absent facilitation: where the organizational environment does not actively support rigorous evidence use, practitioners adapt by engaging the minimum level of evidence-related behavior that satisfies institutional expectations. The qualitative finding that hierarchical interprofessional dynamics function as a systemic disqualifier of nurse-generated evidence extends beyond individual experience to a collective professional phenomenon with structural implications. When nurses learn through repeated professional encounters that evidence-based recommendations will not be acted upon without physician endorsement, the rational professional response is to disinvest from the cognitive and temporal demands of evidence engagement. This is not a failure of professional commitment but a rational adaptation to an institutional environment that does not reward nurse-led inquiry. Rycroft-Malone et al. (2002) identify organizational culture as the strongest predictor of EBP implementation success within the PARIHS framework, and the present findings confirm that hierarchical cultures that position nurses as protocol-followers rather than knowledge-generators represent a fundamental contextual barrier to advancing EBP in Kenyan tertiary hospitals. Addressing this barrier requires not merely building nurses' research capacity but fundamentally restructuring the interprofessional governance of evidence use creating participatory decision-making structures, interprofessional EBP committees, and leadership cultures that recognize and reward evidence-based contributions from nursing staff. The findings generate several concrete implications. First, BScN curricula should be restructured to embed research conduct across the full training trajectory rather than concentrating it in terminal dissertation requirements. Exposure to data collection, analysis practicum, and journal club participation during undergraduate training would normalize research as an ongoing professional activity rather than a graduation ritual (Benner et al., 2010). Second, national referral hospitals should establish protected research time policies, seed funding mechanisms for clinical research, and structured EBP committees and journal clubs evidence-based institutional structures proven to improve evidence culture in resource-limited settings (Fink et al., 2005; Forsman et al., 2012). Third, interprofessional education programs should explicitly address power dynamics between nursing and medicine, cultivating shared governance models for evidence-based clinical decision-making. Fourth, the Nursing Council of Kenya and the Ministry of Health should develop national guidelines linking continuing professional development credits to research conduct and evidence utilization activities, providing career-linked incentives that sustain research engagement beyond formal educational enrollment. This study has several important limitations. The cross-sectional design precludes causal inference regarding the determinants of research involvement and evidence utilization; longitudinal tracking of nurses from graduation through clinical practice would strengthen causal attribution. Self-report data are susceptible to social desirability bias, potentially inflating evidence utilization rates and underreporting barriers. The restriction to two national referral hospitals limits transferability to county, sub-county, or community health settings where workforce demographics, resource environments, and institutional cultures differ substantially. While thematic saturation was achieved within the qualitative sub-sample of 20 participants, the experiences of nurses in less commonly represented departments or specializations may not be fully captured.