| ObjectiveThe purpose of this study is to develop a prediction tool suitable for insertion length of nasogastric tube(NGT)in adult inpatients in China,for improving the accuracy of the insertion NGT length,thus improving the efficiency and nursing quality,and ensuring patient safety and treatment effect.MethodsThis study consists of three sessions:the session 1 is systematic review of predictive factors for NGT insertion length based on literature research;the session 2 is the construction of a prediction models of NGT insertion length based on the crosssectional study;the session 3 is comparative analysis of commonly prediction tools for NGT insertion length based on field survey data.Session 1:Systematic review of predictive factors for NGT insertion length.The PubMed,Embase,Web of Science,CINAHL,CBM,WanFang Data,VIP,and other databases were searched from the inception to January 2022.After the literature screening,the included studies were evaluated and extracted.The included studies involved predictive factors for NGT insertion length.The extracted data were integrated and descriptively analyzed,including the basic characteristics and predictive factors of NGT insertion length to form a questionnaire on the predictive factors.Session 2:Establishment and validation of a prediction tool for NGT insertion length in adult inpatients.A cross-sectional survey with 329 conveniently selected adult inpatients who need to be implanted with NGT from March 2022 to February 2023 was carried to develop a prediction tool in three third-grade A hospital in Lanzhou,and reincluded 52 adult inpatients who need to be implanted with NGT in one of the hospitals for external validation.The survey instument was a self-made questionnaire for predictive factors of NGT insertion length,which was compiled based on the results of systematic review.The data obtained from this survey was input into SPSS 26.0 software for data analysis.All the data were randomly divided into modeling database(230 cases,70%)and internal verification database(cases,30%),and 52 re-collected adult inpatients comprised external validation database.Firstly,the equilibrium among modeling database,internal verification database and external validation database were tested.NGT insertion length prediction model was constructed based on screened predictive factors by multiple stepwise regression.The prediction tool was formed after the thermal map visualization.The area under the ROC curve(AUC)was used to evaluate the discrimination prediction tool for NGT insertion length,and the calibration was tested by Hosmer-Lemeshow and calibration curve graph.Session 3:Comparative analysis of commonly prediction tools for NGT insertion length in adult inpatients.Using external validation data from the session 2,9 prediction tools consisting of what we have established in this study and other 8 tools used at home and abroad were compared for the accuracy.It was defined as NGT accuracy that the difference was within 1.5 cm between predicted insertion length and the actual insertion length and more than 1.5 cm was defined as NGT misplacement.Chi-square test and Fisher’s exact test were used to analyze the inter group differences in the accuracy of different NGT prediction tools.And draw a trend chart between actual insertion length and predicted insertion length of 9 prediction tools.ResultsSession 1:17 articles were included in this systematic review.The results of this systematic review were as follows:1.Characteristics of the included study:The 17 articles included 1990 patients,mainly from 7 countries(China,the United States,the United Kingdom,Belgium,Singapore,New Zealand,and Brazil);9 articles were case-control studies,and the others were cross-sectional studies.2.Risk assessment of bias:The quality evaluation scores of the 17 included articles were all greater than 6 points,indicating modest quality.3.Summary results of predictive factors for NGT insertion length:A total of 19 predictive factors of NGT insertion length were obtained,including gender,consciousness,height,weight,and the length of nose-earlobe,earlobe-umbilicus,earlobe-xiphoid,nose-earlobe-umbilicus,nose-earlobe-xiphoid,earlobe-xiphoidumbilicus,earlobe-xiphoid,glabella-xiphoid,glabella-umbilicus,hairline-xiphoid,hairline-umbilicus,hairline-the midpoint of xiphoid to umbilicus,nose-the midpoint of xiphoid to umbilicus,nose-earlobe-the midpoint of xiphoid to umbilicus,glabella-the midpoint of xiphoid to umbilicus.Session 2:381 participants were included in the cross-sectional study,and the specific results were as follows:1.Descriptive analysis of NGT insertion length and its predictive factors:The modeling database included 230 participants,with the average age of 61.34±15.76 years and 135 male(58.7%).The NGT insertion length was 59.73±2.40 cm;The internal validation database included 99 patients,with the average age of 65.69±16.96 years and 62 male(62.4%).The NGT insertion length was 61.13±2.94 cm;The external validation database included 52 patients,with the average age of 63.12±16.13 years and 33 male(63.5%).The NGT insertion length was 61.15±2.93 cm.The insertion length and 19 predictive factors were balanced and comparable among the modeling database,internal validation database,and external validation database.2.Establishment of prediction tool for NGT insertion length in adult inpatients:The results of multivariate linear stepwise regression analysis showed that it was found that gender(t=-8.717,P<0.001),height(t=2.365,P=0.020),and the length from nose to earlobe then to midpoint of xiphoid to umbilicus(NEXU)(t=5.578,P<0.001)were predictive factors for NGT insertion length.The prediction model was constructed as following:y=0.320*x~1+0.041*x~2-2.795*x~3+34.814y=NGT predictive insertion length,x~1=the length from the nose to earlobe then to the midpoint of xiphoid to the umbilicus,x~2-height,x~3=gender(0=male,1=female).R~2=0.716,and F=112.957(P<0.050).The prediction tool was visualized by performing thermal maps,and named GHNEXU as the tool used gender,height,and the length from nose to ear then to the midpoint of xiphoid to umbilicus.3.Validation of the prediction tool for NGT insertion length in adult inpatients:The prediction tool was validated internally and externally.The AUC respectively were 0.841(95%CI:0.698-0.984)and 0.708(95%CI:0.537-0.879),indicating the good discrimination.Hosmer-Lemeshow results were χ~2=7.567(P=0.385)and χ~2=9.607(P=0.534),and predicted curve was close to actual curve in the calibration curve graph,indicating that the prediction tool had the calibration.Session 3:52 participants were included the comparative analysis of 9 prediction tools for NGT insertion length,and the specific results were as follows:There was statistically different in the accuracy of 9 prediction tools for NGT insertion length(χ~2=156.539,P<0.001).The accuracy in descending order was GHNEXU tool(92.3%)>NEX correction formula(86.5%)>NEX×HX×0.00499+height×0.14615+25.48879 cm(75.0%)>NEX+10 cm(61.5%)>EXU-NE(51.9%)>NEX(46.2%)>Hanson 3 formula(36.5%)>Hanson 1 formula(13.5%)>GWNUF tool(3.8%).ConclusionThe prediction tool for NGT insertion length in adult inpatients is constructed based on gender,height,and NEMU length.And the tool has good discrimination and calibration.The prediction tool is formed by visualizing the model using thermal maps to have high clinical practicality and provide a basis for clinical nurses to implement scientific and accurate NGT insertion. |
PDF Full Text Request