醫療診斷決策模組-以先天性巨腸症為例

Medical diagnosis decision module- example of the Hirschsprung disease

陳安琪(An-Chyi Chen)

指導教授 ： 藍守仁

亞洲大學/健康學院/健康管理研究所/碩士(2005年)

摘要

醫療資源有限情況及注重病人權益下,臨床醫師在臨床上要必須做出對的醫療決策,尤其面對許多種檢查,到底該不該做這是目前最需要且是最重要的課題。所以如果能尋找出一醫療診斷決策模式來提供給臨床醫師,用實證之方法來做醫療決策,不只對全民健保的財務有所幫助且對全民的健康有更好的保障. 故本研究藉由研究先天性巨結腸症(Hirschsprung disease)此一臨床疾病來做進一步探討。
先天性巨腸症(Hirchsprung disease)是一種先天腸胃疾病,每五千個新生兒就有一位患此病,主要病因是其大腸內的神經發育不正常,其常好發生在直腸及乙狀結腸,此病的症狀多元化且常見如,腹脹,腹瀉,便秘,吐,成長不好等且其診斷不是非常容易。尤其下消化道鋇劑攝影檢查之敏感度及特異度並不高,病人如被診斷是此疾病一定要接受開刀,如何確定診斷就非常重要。所以本篇即在探討下消化道鋇劑攝影檢查之敏感度及特異度,並透過成本效益分析方法來提供臨床醫師及放射科醫師做醫療決策.這是一篇回朔性的研究,病患一共有167小兒科病童因臨床上懷疑有可能為巨腸症而列入研究,完成追蹤151位病童,男62位女89位,完成率為89.82%。經過下消化道檢查,直腸黏膜切片及長期追蹤來做最後確診是否為巨腸症,其中有26位被確診為巨腸症。求出下消化道鋇劑攝影檢查之敏感度為0.769特異度為0.464,另透過醫學決策演算過程及邏輯回歸方式,我們求得兩條回歸程式R1-score=0.7717*最大大腸和闔第一腰椎之直徑比+1.3251*過度帶+1.8676*小大腸現象- 0.375*年紀大於2歲, R2score= 0.7791*過度帶+1.2915*過度帶+2.0064*小大腸現象-0.2757*年紀大於2歲-0.5603*24小時鋇劑排出攝影。
利用本研究所演算而得出之接受器操作物性曲線(ROC curve)曲線上之操作點上之似真斜率及先天性巨結腸症之發生率,而去計算出成本及效益比率。因而當知道下消化道鋇劑攝影檢查成本及效益時,這時再結合預期效益分析,就可選定該接受器操作物性曲線適當之操作點。當選定曲線上上之不同點,因而可決定出不同之敏感度及特異度。當放射科醫師將下消化道鋇劑攝影檢查之特徵,帶入上述之公式當分數超過其所導出之門檻值,即可判定為陽性或陰性,也可知道經過這樣的判定會知道該下消化道鋇劑攝影檢查之敏感度及特異度。這樣就可以提供臨床醫師在處理疑似先天性巨結腸症之病患時,能做出對的醫療診斷決策。

關鍵字

先天性巨腸症 ； 下消化道鋇劑攝影檢查 ； 敏感度 ； 特異度 ； 回歸程式 ； 接受器操作物性曲線 ； 醫療決策

並列摘要

Clinician should be done the right medical decision under the limited of medical resource and prospect the patient’s rights and interests on recently. In deed, what’s clinical condition of a patient need to receive a medical examination or test is very difficult to decide it. If we can propose a medical diagnosis decision module under researched by evidence base to provide the clinician to make a right medical decision. It is not only to help the financial affairs of national insurance but also can give a good health system for the people in Taiwan. So the purpose of our research is to propose a medical diagnosis decision module via investigate the Hirschsprung disease. Hirschsprung disease is one kind of congenital gastrointestinal disease and it is very rare. The incidence was about one baby in five thousand live newborns. The disease is a congenital abnormality with an absence of parasympathetic ganglion cells in the affected segment of the rectum, recto-sigmoid, or other colon. The clinical manifestations are variable and common, such as abdominal distended, diarrhea, constipation, vomiting, or growth delay. How to diagnosis the disease is very difficult. Lower the sensitivity and specificity were noted in lower gastrointestinal radiography to diagnose the Hirschsprung disease. So far, the golden treatment for almost patient of Hirschsprung disease is surgery. So we evaluate the specificity and sensitivity of lower gastrointestinal radiograph to diagnosis of Hirschsprung Disease, and provide physician to make medical decision via the method of cost and effectiveness. This paper is a retrospective study. We enrolled 167 patients in this study and completed the questionnaire are 151 patients, those who consist of 62 male and 89 female and the competition rate is 89.92%. Those patients received the clinical diagnosis processing such as lower gastrointestinal radiography, rectal biopsy and long-term clinical follow up, than 26 patients were diagnosed of Hirschsprung disease. The sensitivity of lower gastrointestinal radiography is 0.769 and specificity is 0.464. From processing of medical decision and statistic logistic regression, we got tow formulas of logistic regression. R1-score=0.7717*the ratio of the largest colon’s diameter to the first vertebrae body+1.3251*transitional zone+1.8676*micro-colon - 0.375*age over 2 years old, R2score=0.7791* the ratio of the largest colon’s diameter to the first vertebrae body +1.2915*transitional zone+2.0064*micrcolon-0.2757* age over 2 years old -0.5603*radiography taken 24 hours after the study. We develop ROC curve of lower gastrointestinal radiography, can calculate the specificity and sensitivity from the operating point of the curve that can provide the clinician or radiologist to judge the examination. The likelihood ratio and the prevalence of the Hirschsprung disease are used to calculate the cost the benefit ratio at the operating point. Thus, knowledge of the costs and benefits of the lower gastrointestinal radiography can be used in conjunction with the results of a predictive value analysis to calculate the appropriate operating point for a test. We can get the result of sensitivity and specificity from that operating point. So gastroenterologist and radiologist can got a positive or negative diagnosis when the R1 or R2 value over the value of the threshold which is introduce the characters of lower gastrointestinal radiography into the formula R1 or R2. Additional they can know the sensitivity and the specificity of the lower gastrointestinal radiography when they use this method to diagnose the disease. This medical diagnosis module can provide clinician to do the right medical decision making for those suspect Hirschsprung disease.

並列關鍵字

Hirschsprung ； Lower gastrointestinal radiography ； Sensitivity ； Specificity ； ROC ； Medical decision ； Logistic regression

參考文獻

Walker, Goulet, Klenman, Sherman,Sheider, Sanderson, Pediatric Gastrointestinal Disease Fourth Edition 2004, P570

Baillie C, Kenny S, Williams J,et al. Long term functional outcome and colonic motility following the Duhamel procedure for Hirschsprung’s disease J Pediatr Surg 1999;34:325-30

C.A. Hajivassiliou , Intestinal Obstruction in Neonatal /Pediatric Surgery Seminars in Pediatric Surgery Vol12, No 4,2003, PP241-253

Davidson M, Kugler MM, Bauer CH: Diagnosis and management in children with sever and protracted constipation and obstruction J pediatr 62:261-275,1963

Dobbins WO, Bill AH Jr: Diagnosis of Hirschprung’s disease excluded by rectal suction biopsy, N Eng J Med 272:990-993,1965