Introduction
어느 덧 11월, 펠로우 1년차 수련도 벌써 8개월이 지났습니다
수련 종료 불과 4개월을 앞둔 현재, 성 빈센트 병원 임상강사 선생님들의 대장내시경 실적을 분석합니다
분석 결과가 향후 대장내시경의 질을 향상시키는데 도움이 되었으면 합니다
Methods
Inclusion
- 10월1일~31일 정규시간에 시행된 대장내시경
- 펠로우 1년차 선생님들 판독
Exclusion
- Emergency
- Sigmoidoscopy
- BBPS <6 points
- colon cancer work up
- colon Op한 환자
- No initial photo
Primary outcome
펠로우 개별 insertion time
펠로우 개별 ADR(Adenoma detection rate)
Secondary outcome
Olypmus vs Pentax
Results
Fig2. 펠로우 전체 맹장 삽입 시간 분포
Fig3. 개별 맹장 삽입시간 분포
(A)JIH
(B)KJY
(C)KSH
(D)MJY
(E) OSY
## # A tibble: 5 x 5
## name maximum minimum mean median
## <fct> <dbl> <dbl> <dbl> <dbl>
## 1 JIH 55.1 2.4 12.2 8.3
## 2 KJY 35.6 2 11.7 10.5
## 3 KSH 25.1 2.2 8.31 6.45
## 4 MJY 53.6 1.2 8.78 6.6
## 5 OSY 46.2 2.7 10.2 7.5
Fig4. Boxpolot of insertion time - all fellows
펠로우 선생님들간의 insertion time이 차이가 나는가?
kruskal.test(oct$insertion~oct$name)
##
## Kruskal-Wallis rank sum test
##
## data: oct$insertion by oct$name
## Kruskal-Wallis chi-squared = 15.963, df = 4, p-value = 0.003069
insertion time이 분포는 정규분포를 따르지 않는다.
따라서 kruskal-Wallis 검정을 시행, p-value <0.05 .5명중 적어도 2명이상에서 insertion time차이가 난다
library(nparcomp)
a<-mctp(insertion~name,data=oct)
##
## #----------------Nonparametric Multiple Comparisons for relative effects---------------#
##
## - Alternative Hypothesis: True differences of relative effects are less or equal than 0
## - Estimation Method: Global Pseudo Ranks
## - Type of Contrast : Tukey
## - Confidence Level: 95 %
## - Method = Fisher with 93 DF
##
## #--------------------------------------------------------------------------------------#
##
summary(a)
##
## #----------------Nonparametric Multiple Comparisons for relative effects---------------#
##
## - Alternative Hypothesis: True differences of relative effects are less or equal than 0
## - Estimation Method: Global Pseudo ranks
## - Type of Contrast : Tukey
## - Confidence Level: 95 %
## - Method = Fisher with 93 DF
##
## #--------------------------------------------------------------------------------------#
##
## #----Data Info-------------------------------------------------------------------------#
## Sample Size Effect Lower Upper
## JIH JIH 61 0.5439256 0.4875488 0.5991988
## KJY KJY 42 0.6165215 0.5573298 0.6724501
## KSH KSH 56 0.4215532 0.3671761 0.4779018
## MJY MJY 67 0.4319113 0.3811698 0.4841231
## OSY OSY 60 0.4860884 0.4325518 0.5399461
##
## #----Contrast--------------------------------------------------------------------------#
## JIH KJY KSH MJY OSY
## KJY - JIH -1 1 0 0 0
## KSH - JIH -1 0 1 0 0
## MJY - JIH -1 0 0 1 0
## OSY - JIH -1 0 0 0 1
## KSH - KJY 0 -1 1 0 0
## MJY - KJY 0 -1 0 1 0
## OSY - KJY 0 -1 0 0 1
## MJY - KSH 0 0 -1 1 0
## OSY - KSH 0 0 -1 0 1
## OSY - MJY 0 0 0 -1 1
##
## #----Analysis--------------------------------------------------------------------------#
## Estimator Lower Upper Statistic p.Value
## KJY - JIH 0.073 -0.082 0.224 1.301 0.690624856
## KSH - JIH -0.122 -0.268 0.029 -2.253 0.169612025
## MJY - JIH -0.112 -0.252 0.032 -2.161 0.203434058
## OSY - JIH -0.058 -0.202 0.089 -1.094 0.808980395
## KSH - KJY -0.195 -0.342 -0.039 -3.460 0.007216930
## MJY - KJY -0.185 -0.325 -0.036 -3.440 0.007600161
## OSY - KJY -0.130 -0.277 0.022 -2.387 0.127820783
## MJY - KSH 0.010 -0.132 0.152 0.202 0.999623663
## OSY - KSH 0.065 -0.082 0.208 1.224 0.737196665
## OSY - MJY 0.054 -0.085 0.191 1.081 0.815615294
##
## #----Overall---------------------------------------------------------------------------#
## Quantile p.Value
## 1 2.780687 0.00721693
##
## #--------------------------------------------------------------------------------------#
Fig5. 나이에 따른 insertino time의 분포
Fig6. ADR (Adenoma Detection Rate)
GI Endoscopy
(A) 전체 펠로우 ADR
제외: Polypectomy 처방을 받고 검사, 조직검사 결과가 안나온 경우
Table2. 개별 펠로우 회수 시간
polypectomy, biopsy등 술기 없이 순수히 관찰만하고 나온 경우로 제한
모두 평균 6분 이상씩 관찰하고 있음
Fig7. Olympus vs Pentax
(A) Olympus 와 pentax insertion time 비교
##
## Wilcoxon rank sum test with continuity correction
##
## data: olympus$insertion and pentax$insertion
## W = 7708, p-value = 0.08702
## alternative hypothesis: true location shift is not equal to 0
Olympus가 1분정도 빨리 들어가는데,통계적 유의성은 없다.
(B) Olympus vs pentax ADR 비교
##
## 2-sample test for equality of proportions with continuity
## correction
##
## data: table(adr1$scope, adr1$adenoma)
## X-squared = 4.8125, df = 1, p-value = 0.02825
## alternative hypothesis: two.sided
## 95 percent confidence interval:
## -0.28916502 -0.02252329
## sample estimates:
## prop 1 prop 2
## 0.5974026 0.7532468
ADR의 로지스틱 회귀분석
##
## Call:
## glm(formula = adenoma1 ~ sex + age + scope + name, family = binomial,
## data = adr1)
##
## Deviance Residuals:
## Min 1Q Median 3Q Max
## -1.4263 -0.9313 -0.5671 1.1334 2.3176
##
## Coefficients:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) -2.78700 0.79108 -3.523 0.000427 ***
## sexM -0.02940 0.29710 -0.099 0.921183
## age 0.04524 0.01158 3.908 9.32e-05 ***
## scopepentax -0.64906 0.33128 -1.959 0.050084 .
## nameKJY -1.04789 0.52120 -2.011 0.044376 *
## nameKSH -0.58161 0.45249 -1.285 0.198668
## nameMJY 0.03663 0.42857 0.085 0.931887
## nameOSY -0.46269 0.44550 -1.039 0.298993
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for binomial family taken to be 1)
##
## Null deviance: 299.31 on 230 degrees of freedom
## Residual deviance: 268.67 on 223 degrees of freedom
## AIC: 284.67
##
## Number of Fisher Scoring iterations: 4
Summary
1. 우리 병원 median insertion time 7.75분 이다
3. Pentax 쓴다고 insertion time이 길어지지 않는다. 하지만 ADR은 떨어지는 경향이 있다.