穴位贴敷疗法对失眠障碍患者PSQI改善率临床预测模型的建立与验证OA

Objective:To construct a clinical prediction model for the impact of acupoint patch therapy on the improvement rate of PSQI in patients with insomnia disorder(ID),providing insights and methods for predicting the outcomes of insomnia in ID patients treated with acupoint patch therapy.Methods:A retrospective cohort study was conducted,collecting clinical data of 64 ID patients who underwent acupoint patch therapy at two centers,the Acupuncture Hospital of China Academy of Chinese Medical Sciences and Fengtai Times Community Health Service Station,from June 2019 to June 2020 at 0 weeks,1 week,2 weeks,3 weeks,4 weeks,and 6 weeks(2 cases excluded,a total of 372 data).Single-factor Logistic regres-sion analysis,forward and backward stepwise regression analysis were used to select variables for building the nomogram prediction model.The opti-mal model was selected based on the Akaike information criterion(AIC).The factors of the optimal model were further used to construct a nomo-gram prediction model using the"rms"package in the R language.The discriminative ability of the model was assessed by the area under the receiver operating characteristic curve(AUC),calibration curve,and the Hosmer-Lemeshow goodness-of-fit test evaluated the calibration.Internal validation of the model was performed using the"Bootstrap resampling method".Results:Treatment time,duration of illness≤12 months,12 months<duration of illness≤36 months,fatigue level,previous use of sleep-promoting drugs,general health status,social function,physical pain,liver depression transforming into fire combination with AIC of 300.31 was the optimal model.The variance inflation factor(VIF)was all<5,and the tolerance was all>0.5,indicating a small possibility of multicollinearity.These factors were independent predictors of the acupoint patch therapy for insomnia out-comes in ID patients.The AUC of the ROC curve for the prediction model was[0.856,95%CI(0.814,0.898)],indicating a good discriminative abil-ity.The AUC of each variable predicting the outcome event alone ranged from 0.5 to 0.8,all lower than the AUC value of the joint prediction of the outcome event,indicating that joint prediction had better discriminative ability and predictive power.The calibration curve showed that the fitting curve of the prediction model was close to the ideal curve,indicating good calibration.The Brier score was[0.125,95%CI(0.105,0.145)],indicat-ing good overall predictive performance of the model.The Hosmer-Lemeshow goodness-of-fit test showed P=0.152(P>0.05),suggesting a good fit of the model.Conclusion:The clinical prediction model constructed with treatment time,duration of illness≤12 months,12 months<duration of ill-ness≤36 months,fatigue level,previous use of sleep-promoting drugs,general health status,social function,physical pain,liver depression trans-forming into fire has good predictive efficacy.It can provide certain reference value for clinical decision-making in the treatment of insomnia disor-der with acupoint patch therapy.