<![CDATA[The digitization of clinical workflows through Electronic Health Records (EHR) is a global imperative aimed at enhancing data accuracy and care coordination. However, in resource-constrained environments, the transition from paper-based systems to digital platforms often surpasses the readiness of existing infrastructure. While systems such as the Lightwave Health Information Management System (LHIMS) in Ghana offer the promise of increased efficiency, they also introduce critical dependencies on unstable power and internet connectivity. This situation creates a "Digital Efficiency Paradox," wherein the urgency to document data swiftly before a potential power outage inadvertently diminishes the quality of clinician-patient interactions. This study employs a qualitative-driven process modeling approach at Juaben Municipal Hospital (N=10). We utilize formal Business Process Model and Notation (BPMN 2.0) semantics to reconstruct clinical workflows and apply the Control-Flow Complexity (CFC) metric to quantify the cognitive load shift from manual ($W_{\text{pre}}$) to digital ($W_{\text{post}}$) systems. Computational analysis reveals that while LHIMS reduced patient retrieval latency by approximately 96%, it increased structural complexity (CFC) from 3.0 to 14.0, thereby imposing a higher cognitive burden. Crucially, we identified a phenomenon of "Infrastructure-Induced Process Deadlock," where power outages result in total system paralysis ($\mathcal{I}(\tau)=0$), compelling clinicians to resort to risky hybrid workarounds. Paradoxically, the anxiety of potential system failure drives staff to prioritize "screen time" over "care time," creating a tunnel vision effect. The study challenges the "always-online" paradigm in the Global South. We conclude that digital efficiency must be balanced with structural resilience, advocating for an "Offline-First" architecture that decouples clinical documentation from grid instability to preserve the human element of care.]]>
