View : 1581 Download: 0

입원환자의 영양불량 위험 평가를 위한 영양검색도구 개발

Title
입원환자의 영양불량 위험 평가를 위한 영양검색도구 개발
Other Titles
Nutritional screening tool development for inpatient assessment of risk of malnutrition
Authors
김미향
Issue Date
2013
Department/Major
임상보건과학대학원 임상영양학전공
Publisher
이화여자대학교 임상보건과학대학원
Degree
Master
Advisors
조미숙
Abstract
The prevalence of malnutrition in hospitalized patients is different according to researchers, research methods and malnutrition criteria, but the inpatient who is under high malnutrition risk has higher complications and a incidence of death and longer hospital stay, which leads to an increased medical costs. However, 25% of inpatients have been reported as treatable malnutrition status, so early screening is required to the hospitalized patients who are at risk of malnutrition from the beginning of the hospitalization and to provide adequate nutritional therapy. However, unlike the other foreign countries, dietitians have a lot of difficulties in conducting direct nutritional assessment for all assigned inpatients in terms of time, considering human resources, cost and time because of other management and works except clinical nutrition. Therefore, to reduce time, human resource and cost that could be spent in case of evaluating all inpatients by direct interview, the urgent development of nutrition screening tool which is simple, fast and precise is required. Therefore this study tried to develop a nutritional screening tool which could examine high risk malnutrition patients in a simple, fast and cost-effectively way by utilizing index such as biochemical data, anthropometry data, diet history, diagnosis which were proven to have high orrelation in qualified nutrition search method of domestic and foreign countries and also commonly available at the beginning of the hospitalization, by analyzing the correlation between nutritional evaluation result by clinical dietitian. The study investigated number of 1,901 age of 18 hospitalized patients in tertiary university medical center in Gyeonggi-province from 15th September, 2010 to 15th October, 2010. After collecting informations about gender, age, anthropometric data, weight change, 8 biochemical data (Albumin, Hemoglobin, Hematocrit, Total lymphocyte count, Blood urea nitrogen, Creatinine, Cholesterol, C-reactive protein), 10 diagnosis(diabetes, renal failure, liver cirrhosis, hepatic encephalopathy, Congenital metabolic diseases, sore, multiple trauma, burn> 10%, AIDS, Tuberculosis), meal pattern, 9 nutrition-related symptoms(nausea, vomiting, constipation, diarrhea, dysphagia, chewing difficulty, poor oral intake, NPO> 3days, tube feeding), appetite, intake, nutrition requirement through nutritional assessment resulted by direct dietitian interview and EMR(Electronic Medical Record), and the study used the SPSS 18.0 statistical program to present the data. After analyzing nutritional assessment results through an experienced dietitian's direct interview of all hospitalized patients within 48 hours, the study showed that among the 1,908 patients, the initial adequately nourished group included 1,263(66.4%) patients, 638 patients(33.6%) were estimated over mild malnutrition status, and severe malnutrition group included 21 patients(1.1%). To analyze the correlation between nutrition-related indexes collected through EMR and nutritional assessment result evaluated by an experienced dietitian, the study select the final nutritional screening indicators after deleting the low correlation indicators by logistic regression analysis and backward Wald method. The study took statistics each nutritional screening indexes after grading it by following case; if index is less than the reference value, set 0, if index is greater than the reference value, set 1. After the study set the dependent variable as initial nutritional assessment and classified adequate status as good nutrition group and over mild malnutrition status as poor nutrition group. The final regression equation was calculated except in the case the constant term was negative number and P-value was less than 0.05; 2.5×Body mass index + 2×Weight loss + 1.5×albumin + 1.5 ×hemoglobin + 0.5×Blood urea nitrogen + 0.5×C-reactive protein +3×Liver cirrhosis + 2.5×Dysphagia + 1.5×Chewing difficulty + 2×Poor oral intake). By the ROC curve analysis, the benchmark was setted as 2.25 which was the highest value in Youden index, and sensitivity was 84.9%, specificity was 81.3%. To utilize the drawn regression equation as new nutritional screening tool, the study executed kappa test to verify conformity degree between the existing nutrition screening method and nutrition evaluation result by dietitian interview. The κ value by the test was 0.698, which meant the new tool was substantial with nutrition evaluation result. The newly developed nutritional screening tool compound objective index and subjective index to supplement the limitation of each one. In addition, the tool is considered "effective" in terms of human resource and cost, because it is possible to utilize by activating common blood test index and EMR datas which are attained from early hospitalization.;입원환자의 영양 불량 유병률은 연구방법과 연구자, 영양불량 기준에 따라 차이가 있으나 입원 초기 영양상태가 불량한 환자군에서 영양상태가 양호한 환자군에 비해 재원기간이 길어지고, 합병증, 사망률이 높고 이로 인해 의료비용이 상승되는 것으로 보고되고 있다. 따라서 입원 초기에 영양불량 위험환자를 조기에 선별하여 적절한 영양치료를 제공하는 것이 필요하다. 본 연구에서는 경기도 소재의 1개 3차 대학병원에 2010년 9월 15일부터 10월 15일 까지 입원한 18세 이상 성인 환자를 대상으로 전자의무기록 분석을 통한 후향적인 연구방법으로 입원 48시간 이내 숙련된 임상영양사의 환자 대면면담에 의해 시행된 영양평가결과와 의무기록을 통해 수집한 영양관련 지표들과의 상관성을 분석하여 입원 초기에 통상적으로 이용 가능한 지표들을 활용하여 간단하고 빠르며 비용 효과적으로 영양불량위험 환자를 판별하기 위한 영양검색도구를 개발하고자 하였다. 영양검색 지표는 기준값 이하와 이상으로 구분하여 점수화 한 후 SPSS 18.0프로그램을 이용하여 통계 처리하였다. 연구대상자 1,901명중 남자는 1,003명(52.8%), 여자는 898명(47.2%) 이었으며, 평균연령은 55.13 ± 16.41세였고 65세 이상 노인은 613명으로 32.2%를 차지하였다. 평균 BMI는 23.16 ± 7.01kg/m2으로 나타났다. 임상영양사의 대면면담을 통한 영양평가 결과 1,263명(66.4%)는 입원 초기 영양상태가 양호한 것으로 평가되었으며, 638명(33.6%)은 mild malnutrition 이상의 영양불량을 가진 것으로 평가되었다. 연령에 따른 영양 불량율을 비교한 결과 연령이 증가할수록 영양불량율이 유의하게 증가하는 것으로 나타났으며(p<0.001), 연령이 10세 증가할수록 영양불량 odds ratio는 1.35배 증가하였다. 연구대상자의 영양 상태에 따른 임상결과를 비교한 결과 평균 재원일수는 11.33 ± 14.82일 이었으며, 영양상태가 양호한 환자(9.52 ± 1.33일)에 비해 영양불량이 있는 환자(14.9 ± 19.55일)에서 통계적으로 유의하게 길었다(p<0.001). 합병증, 임상경과가 나쁨, 사망률은 영양상태 양호군보다 영양 불량군에서 각각 5.07, 4.02, 7.72배 증가하였다. 이분형 로지스틱, 후진 wald 방법을 사용하여 통계 분석한 결과 마이너스값인 상수항과 p-value가 0.05 미만인 지표를 제외하고 최종 회귀식(2.5×BMI + 2×Weight loss + 1.5×albumin + 1.5×hemoglobin + 0.5×BUN + 0.5×CRP + 3×Liver cirrhosis + 2.5×Dysphagia + 1.5×Chewing difficulty + 2×Poor oral intake)을 정하였다. ROC 곡선 분석을 통해 Youden index 값이 가장 높은 2.25를 영양불량 기준점으로 설정하였으며, 민감도는 84.9%, 특이도는 81.3%였다. 새로 개발한 영양검색 도구는 객관적인 지표와 주관적인 지표를 혼합 사용하여 각각의 지표가 가진 제한점을 보완하도록 하였으며, 환자의 입원 초기 간호사가 수행하는 입력된 전자 의무기록 정보들과 일반적인 혈액검사 지표를 활용하여 영양 검색을 수행할 수 있어서, 임상 영양 치료측면에서 입원 환자의 영양관리를 위한 인력과 비용, 효과 측면에서 효율적인 검색 도구가 될 것으로 본다.
Fulltext
Show the fulltext
Appears in Collections:
임상보건융합대학원 > 임상영양학전공 > Theses_Master
Files in This Item:
There are no files associated with this item.
Export
RIS (EndNote)
XLS (Excel)
XML


qrcode

BROWSE