표제지
국문초록
목차
약어목차 15
제1장 서론 17
제1절 NSAID 사용의 개요 17
1. NSAID의 통증 치료 17
2. NSAID의 작용기전과 부작용 18
제2절 NSAID 유발성 위장관 합병증 22
1. NSAID 관련 위장관 합병증 발생의 위험인자 23
2. 위장관 합병증 고위험 환자의 분류 25
3. NSAID 사용 고위험 환자의 위장관 합병증 예방 26
제3절 NSAID 유발성 위장관 합병증 관련 기존 연구 29
1. NSAID 사용 고위험 환자의 위장관 합병증 연구 29
2. NSAID 유발성 위장관 합병증 위험예측모델 연구 30
제4절 노인에서 위장관 합병증의 취약성 34
1. 노인환자의 높은 NSAID 사용빈도 34
2. 노인환자의 심각한 위장관 합병증 증가 34
제5절 NSAID 관련 위장관 합병증 예측의 필요성 36
제6절 연구의 목적 38
제2장 NSAID와 항응고제 병용 환자의 심각한 위장관 합병증 위험과 예방 39
제1절 연구 방법 39
1. 자료원 39
2. 연구 설계 및 대상 환자 40
3. 결과변수 42
4. 약물 사용의 평가(exposure) 46
5. 공변량 52
6. 위장관 합병증 예방요법 비교 60
7. 통계 분석 61
제2절 연구 결과 62
1. 연구 코호트의 기본 특성 62
2. 심각한 위장관 합병증의 발생과 위험 67
3. 심각한 위장관 합병증 예방요법 비교 72
4. 민감도 분석 77
5. 하위그룹 분석 81
제3장 위장관 합병증 위험예측점수 모델의 개발과 검증 94
제1절 연구 방법 94
1. 자료원 94
2. 연구 모집단 95
3. 사례군과 대조군의 정의 97
4. 후보 예측인자 99
5. 누락 데이터 109
6. 통계 분석 109
제2절 연구 결과 117
1. 연구 대상 인구와 기본 특성 117
2. 예측모델의 개발 123
3. 예측모델의 성능 146
4. 고위험 환자의 예측 150
제4장 고찰 및 결론 154
제1절 연구의 의의 154
제2절 연구의 한계 164
1. 심각한 위장관 합병증 발생의 확인 164
2. 건강보험 청구자료에서 변수 식별의 한계 166
3. 예측모델의 개발, 검증 연구의 대상자의 불일치 167
4. 외부 검증의 한계 168
제3절 결론 170
참고문헌 171
Abstract 186
Table 1. Risk factors for NSAID-induced GI complications according to guidelines in Korea, the United States, and Japan 24
Table 2. Strategies for preventing NSAID-induced GI complications according to guidelines in Korea, the United States, and Japan 28
Table 3. Comparison of previous prediction models related to NSAID-induced GI complications 32
Table 4. Potential predictive risk factors from previous studies related to NSAID-induced GI complications 33
Table 5. Diagnosis for serious GI complications 44
Table 6. Exposure medications and ATC codes used in Study 1 48
Table 7. Diagnosis and ICD codes used in Study 53
Table 8. Medications as covariates and their ATC codes used in Study 1 58
Table 9. Baseline characteristics of the population in Study 1 65
Table 10. Incidence of serious GI complication in Study 1 68
Table 11. Incidence rates of serious GI complications among patients who started anticoagulants and/or NSAIDs 69
Table 12. Risk of serious GI complications among patients who started anticoagulants and/or NSAIDs 71
Table 13. Comparison between the group receiving GI prophylaxis and no prophylaxis for serious GI complications undergoing concomitant NSAIDs and anticoagulant therapy 73
Table 14. Comparison the risk of serious GI complications between GI prevention strategies in patients receiving concomitant therapy with NSAIDs and anticoagulants 75
Table 15. First sensitivity analysis in Study 1: alternative definitions of overlap between concomitant medications 78
Table 16. Second sensitivity analysis in Study 1: without adding 25% period to exposure 80
Table 17. Risk comparison of preventive strategies for GI complications in concurrent use of NSAIDs and anticoagulants, stratified by anticoagulant type... 82
Table 18. List of potential predictive variables in study 2 100
Table 19. Medications and ATC codes used in Study 2 104
Table 20. Comparison of ridge, LASSO, and Elastic Net regression 112
Table 21. Baseline characteristics of derivation and validation cohort in Study 2 120
Table 22. Odds ratios of serious GI complication in the derivation cohort 124
Table 23. Risk scores derived from multivariate logistic regression analysis 131
Table 24. Performance of prediction scores using logistic regression analysis 134
Table 25. Risk scores in developed prediction model 139
Table 26. Comparison of risk scores calculated using multivariate logistic and LASSO regression analysis 142
Table 27. Performance of prediction score by derivation and validation cohorts 147
Table 28. Accuracy of the risk prediction model using the cut-off values 151
Table 29. Ten cases with high frequency predicted as high-risk patients from external validation dataset 152
Figure 1. Prostaglandin synthesis through cyclooxygenase Adapted from Wolfe MM et al. Engl J Med. (1999)... 20
Figure 2. Diagram of the design and protocol in Study 1 Since both anticoagulants and NSAIDs were used, cohort entry date... 41
Figure 3. Construction of admission episodes The admission episode was linked to the start dates and the end... 43
Figure 4. Definition of continuous medication use The duration of medication exposure included an additional 25% of... 47
Figure 5. Constructing data for time-dependent variables Time intervals were divided for each change in NSAID exposure... 51
Figure 6. Flowchart of the cohort selection for Study 1 Abbreviations: GI, gastrointestinal; NSAID, nonsteroidal... 63
Figure 7. Forest plot comparing preventive strategies for GI complications in concurrent NSAIDs and anticoagulants therapy... 76
Figure 8. Diagram of the design and protocol in Study 2 Abbreviations: NSAID, nonsteroidal anti-inflammatory drug; GI, gastrointestinal. 98
Figure 9. Flowchart of population selection in Study 2 118
Figure 10. Calibration curves for serious GI complication in the risk scores using logistic regression analysis... 135
Figure 11. Selection of parameters at varying lambda values (A) Tuning parameter selection in the LASSO model used 10-fold... 137
Figure 12. Calibration curves of serious GI complication in final prediction scores... 148
Figure 13. Calibration plots by deciles of NSAID users The risk of GI complications increases as the prediction scores increase.... 149