원문정보
초록
영어
Digital efficiency in high-risk care settings can inadvertently silence life-critical signals embedded in unstructured narratives. We conceptualize Silent Risk as a governance failure in which clinically consequential cues (e.g., fall precursors, acute deterioration warnings, self-harm threats) are context-stripped and absorbed into routine administrative labels. Using 870 narrative records accumulated over 57 months in a long-term care facility, we triangulated iterative qualitative coding (NVivo) with cross-tabulation to trace concealment pathways within the legacy five-category scheme (Physical, Cognitive, Emotional, Behavioral, Others) and to operationalize concealment as the Risk Silence Rate (SR)—the proportion of records containing HRO-critical risk cues absorbed into non-critical categories. In an in-depth adjudication subset (N = 25) purposively sampled from screened potential risk cases, 12/25 contained Life_Safety risks dispersed across Physical, Emotional, and Cognitive categories; four recurrent Silent Risk types emerged: Life_Safety, Dignity, Depression, and Resistance (Meaningful Refusal). Grounded in the High Reliability Organization principle of preoccupation with failure, we propose a Responsibility-Embedded VG-HITL governance architecture that preserves context via a Value Graph, routes high-risk/uncertain/conflicting cases into mandatory expert adjudication, and feeds corrections back via RLHF to update detection rules and judgment logic, reframing decision quality toward minimizing fatal omissions and strengthening ethical accountability.
한국어
돌봄 현장의 디지털 효율화는 비정형 서술 기록의 맥락을 평탄화하여 낙상 전조ㆍ자해 위협 등 생명안전 신호를 일반 범주로 흡수ㆍ은폐하는 ‘침묵의 위험(Silent Risk)’을 초래한다. 본 연구는 요양기관 서술 기록 870건(57개월 누적, 무작위 표본)을 NVivo 기반 정성코딩과 교차표 분석으로 삼각검증하여, 기존 5분류(신체ㆍ인지ㆍ정서ㆍ행동ㆍ기타) 체계에서 고위험 단서가 어떻게 분산․누락되는지의 은폐 패턴을 규명하고 이를 ‘침묵률(Risk Silence Rate, SR)’로 정량화했다. 심층 검토 사례(N = 25)에서 48%(12/25)가 Life_Safety 위험임에도 신체ㆍ정서ㆍ인지 범주로 분산 은폐되었으며, 추가로 존엄(Dignity)ㆍ우울(Depression)ㆍ의미 있는 거부(Resistance)의 4개 핵심 위험 유형이 도출되었다. 이를 바탕으로 HRO의 ‘실패에 대한 집착’ 원리를 구현한 책임 내재형 VG-HITL 아키텍처를 제안한다. 제안 모델은 가치-그래프(Value Graph)로 맥락 단서를 구조화하고, 고위험 신호ㆍ불확실성ㆍ분류 충돌 사례를 전문가 재판정의 필수 경로로 자동 큐잉한 뒤, 교정 피드백을 RLHF로 환류시켜 ‘놓친 실패’의 재발을 줄이도록 설계된다. 본 연구는 AI 의사결정 평가를 정확도 중심에서 치명 위험 누락 최소화와 윤리적 책임성(생명안전ㆍ존엄ㆍ우울ㆍ거부 위험의 가시화) 중심으로 재구성함으로써 고위험 돌봄 환경의 책임 기반 거버넌스 설계 원리를 제시한다.
목차
Ⅰ. Introduction
1.1 Research Background and Problem Statement
1.2 Research Purpose and Differentiation
Ⅱ. Theoretical Background
2.1 High Reliability Organizations (HRO) and Preoccupation with Failure
2.2 Context Stripping of Unstructured Data and Silent Risk
2.3 Proposed Conceptual Definition and Distinction of Silent Risk
2.4 Responsibility-Embedded Governance and VG-HITL Architecture
Ⅲ. Research Methodology
3.1 Research Framework: Methodological Triangulation
3.2 Data Collection and Sampling
3.3 Analysis Procedure
Ⅳ. Research Results
4.1 Distribution Characteristics of Analysis Data
4.2 Typology of ‘Silent Risk’ through Qualitative Coding
4.3 Analysis of Concealment Patterns of Silent Risk
4.4 Deep Contextual Analysis of Discrepancy Cases
4.5 Summary of Analysis Results
Ⅴ. Conclusion and Implications
5.1 Summary of Research Results
5.2 Academic and Practical Implications
5.3 Research Limitations and Suggestions for Future Research
References
Abstract
