Predicting the loss to follow‐up (LTFU) of HIV/AIDS patients in China using a recency‐frequency (RF) model

Abstract

Background We constructed a recency‐frequency (RF) model for predicting the loss to follow‐up (LTFU) in HIV/AIDS patients in China. Methods Data on HIV/AIDS outpatients in the research unit from 1 August 2009 to 30 September 2020 and from 1 October to 31 December 2020 were exported as the observation and prediction datasets, respectively. The classic recency‐frequency‐monetary (RFM) model was expanded into RFm, RF, RFL and RFmL models. In the observation dataset, the best predictive model was obtained using k‐means clustering and C5.0 verification. Then, two rounds of k‐means modelling were performed on the best model: data with R ≤ 6 months were retained, randomly divided into a training set (70%) and a testing set (30%) and used to perform the second round of modelling to subdivide the churn and non‐churn patients. Next, an ANN algorithm was used to predict LTFU, and the confusion matrix with prediction datasets was constructed. Results The observation and prediction datasets included 16 949 and 10 748 samples, respectively. The RF model with three clusters and a quality of 0.82 was the best predictive model. From the observation set, 13 799 samples were retained, and the model accuracy was 100% on the training and testing sets. These 13 799 samples were subdivided into 1563 samples of churn patients and 12 216 samples of non‐churn patients. The accuracy of ANN prediction was 99.89%. The accuracy and precision of the confusion matrix were 85.41% and 99.76%, respectively. Conclusion The RF model is effective in predicting the LTFU of HIV/AIDS patients in China and preventing its occurrence.

Subdividing ART patients and analyzing the medical burden by modeling of CD4 cell count

“…As described in the literature,[ 3 , 19 20 21 22 23 ] the data of HIV/AIDS patients visiting the outpatient clinic from August 2009 to December 2020 were exported from the HIS of the only grade A tertiary hospital in Shanghai, China, that is designated to treat AIDS patients. The extracted data included the patient’s identity number (ID No.…”

“…Following literature,[ 19 20 21 22 23 , 25 ] patients were distinguished through RF modeling using the k-means clustering algorithm, and the model was evaluated using the C5.0 decision tree. The data of the ART-treated patients were preserved [ Table 1 ].…”

“…The 15 types of ART-treated patients were subdivided using the k-means modeling of the previously obtained important variables, as described in literature,[ 19 20 21 22 23 ] and the model was validated using the C5.0 decision tree. Moreover, we obtained the major predictive variables [ Tables 1 and 3 ].…”

“…Ten new variables were generated according to the calculation formula described in Tables 3 and 4 , including the cycle of ART (years), frequency of visits (visits/person/year), percentage of examination expenditure (%), ART-MB (yuan/person/year), reducible ART expense (yuan/person/year), exceeded frequency of visits (visits/person/year), reducible medical expense per year (yuan), reducible fees for two examinations per year (yuan), payable medical expense of ART (yuan/person/year), and percentage of reducible expense (%). According to literature,[ 19 20 21 22 23 , 28 ] the features and medical burden were analyzed for different types of ART-treated patients [ Tables 3 , 4 and Figure 1 ].…”

“…In this work, we extracted the valid data of HIV/AIDS patients visiting the outpatient clinic of the only grade A tertiary hospital in Shanghai, China, that is designated to treat AIDS patients, between August 2009 and December 2020, from the hospital information system (HIS). We divided ART patients (non-churn patients) based on the recency-frequency (RF) model,[ 19 20 21 22 23 ] analyzed the medical burden of the different types of ART patients, and discussed how to reduce ART costs, ultimately reducing the medical burden of ART (ART-MB), which may be used as a reference for ART-treated patients in various countries and regions. [ 24 ] The CD4 cell count is an important variable to subdivide ART-treated patients.…”