Chronic heart failure (CHF), a syndrome induced by cardiac dysfunction, is a leading cause of cardiovascular disease-related deaths. Approximately 26 million people suffer from CHF worldwide, and the number has been growing in recent years. Furthermore, CHF declines the quality of life of CHF patients and places increasingly heavy burdens on the affected families and the entire society (1–3). It has become a major threat to social progress and human health (4). Therefore, early assessment and management of patients with CHF are critical in primary care.

With the advances in diagnosis and treatment techniques in recent years, the survival of CHF patients has improved (5). However, only the cross-section data were considered in most classification criteria, while the longitudinal changes in the patients were neglected (6). The prolonged survival of HF patients could facilitate state shifts in assessing treatment effects. However, these state shifts were ignored in many studies (7, 8), leading to patients’ worsened condition and decreased quality of life (9).

Patient-reported outcomes measurements (PROMs) filled out by the patients accurately reflect their experiences and are important for clinical diagnoses and treatment. PROMs provide the indices of disease morbidity and prognostic assessment in clinical research and practice, thus having received wide attention (10). Tian et al. (11) devised PROMs for Chinese patients with CHF, which could fully reflect the state of the patients in the physiological, psychological, social, and therapeutic domains. CHF-PROMs facilitate the identification of different patient subgroups and the creation of longitudinal cohorts to analyze response shifts in different subgroups. However, the patient’s condition is a latent variable that cannot be directly observed. Therefore, latent transition modeling is required for subgroup classification and state transition analysis of CHF-PROMs data.

This paper aims to (1) explore the longitudinal dynamic development trend of CHF patient subtypes by analyzing the subtype characteristics of CHF patients and their disease conditions over time; (2) analyze the predictive effect of CHF patient subtype changes over time, gender, age, etc., on the change of subtype category.

In this study, 712 questionnaires were distributed, of which 504 included two follow-up visits, and 497 were valid, with an effective rate of 98.6% (Figure 1). Among the 497 patients, the mean age was 67 ± 14.4 years, of which 280 (56.34%) were male. In this study, an unrotated principal component analysis was conducted using Harman’s single factor test on data from three time points to test the common method variance. The results indicated no excessive explanation of any of the factors for the data at three time points, suggesting no significant common method bias at three time points.

In this study, the transition of latent subtypes over time in CHF patients and its influencing factors were investigated based on the distribution of latent subtypes in CHF patients. LTA was adopted to study the state transition in the patients in the physiological, psychological, social, and therapeutic domains.

Patients with CHF were classified according to latent profile analysis (LPA), LPA-based LTA, and patient-reported outcomes. Specifically, the patients were classified into high- and low-level groups in the physiological, psychological, and social domains according to their subdomain scores. For instance, in the physiological domain, patients in the high-level group have a relatively better somatic condition, appetite and sleep, and physical independence, while patients in the low-level group have a relatively lower performance in the three subdomains. Patients were also classified into two groups in the therapeutical domain. The subdomain scores show that patients with high satisfaction and adherence are prone to drug side effects or to worry about the development of drug side effects. Studies on other diseases, such as psychiatric disorders and hypertension, have shown that medication side effects may affect patient adherence and satisfaction (20–22).

Latent transition analysis reveals that in the physiological domain, the high-level group is relatively stable in both phases, and the low-level group is more prone to state transition in the first phase and tends to stabilize in the second phase. Some studies on CHF focused on the effect of multiple factors on cardiac function classification or the prognosis of different cardiac function classes and ejection fractions but rarely touched on the physiological effect stratification (23–25). Patient state transition analysis in this study shows that patient states increase steadily before increasing at a decreasing rate. Therefore, additional attention should be given to patients with a transition from high to low levels. Multivariate analysis shows that gender, age, tea consumption, smoking, alcohol consumption, physical activity, and light diet statistically affect the patients. Female patients are more likely to transition to the low-level group in the physiological domain after discharge. Therefore, more attention should be paid to the prognostic changes in women’s physical states. Age is an important influencing factor in patient state. High age has been shown as a high-risk factor for poor prognosis in patients with acute heart failure (26), and extended care for elderly patients is a way to enhance the prognosis of patients (27).

In the psychological domain, most patients’ psychological level was improving, and the scores of the two groups were getting closer. The results indicate that although the patients’ states have changed, their overall psychological level is gradually improving. The change in patients’ psychological levels is similar to that in patients’ physical levels, probably because patients’ psychological levels are influenced by physiological changes (28). The probability and extent of state transition in mental health are greater than those in physical health, and the score analysis shows that an increase in patient scores cannot fully identify an improvement in the patient’s mental state (29). As mental health is a part of overall health, improving the patient’s psychological level helps to improve the patient’s physical condition (30). With the consideration of the normal psychological changes in patients and the actual scores, patient states should increase steadily before increasing at a decreasing rate. Therefore, patients transitioning from the high-level group to the low-level group (deteriorating patients) should be given additional attention. Analysis of psychological risk factors suggests that the statistically significant influencing factors are gender, occupation, smoking, alcohol consumption, and physical activity. Like the physiological domain, female patients are more prone to deterioration (31). Smoking is a risk factor for psychological state shifts in patients (32). Alcohol consumption has been shown to affect the patient’s psychological level positively. The conclusion is contrary to previous studies probably because moderate alcohol consumption can alleviate patients’ anxiety, and different levels of alcohol consumption have different effects on patients’ psychological and physical conditions. This study also suggests that physical exercise positively affects the patient’s psychological level, and prolonged exercise may be more effective in stabilizing patients’ emotional and psychological states. Different prognostic approaches should be taken to improve the psychological level of CHF patients at different stages and psychological levels.

In the social domain, patients’ social support and support utilization tend to be stable after discharge and begin to decline 1 month after discharge, which may be related to the improvement from the disease. However, decreased support does not mean reduced care for the patients. More attention should be paid to patients transitioning from the low- to the high-level groups (improving patients) to identify the factors for improvement and help patients with deteriorated and unimproved condition. Analysis of patient-specific changes in social levels shows that age is the only factor associated with changes in social levels, but the effect is small. Patients’ social support levels and support utilization correlate with patients’ psychological and physical states, and studies on other diseases suggest the influence of gender and age on the state shifts in the social domain (33).

In the therapeutical domain, the change in patients’ treatment experience levels is similar to that in the social domain, possibly due to the decreased adherence and satisfaction with the physician after discharge compared with the first discharge or the relaxed concern of the physician on patient’s treatment experience after discharge. Improved care, patient satisfaction, and adherence positively affect the patient (31). Therefore, patient management should be extended for a long period after discharge. Education level and income are found to positively affect the study of factors influencing changes in patients’ treatment experience. The reason for the result may be that patients with higher incomes and education levels are more likely to trust doctors and have sufficient financial resources to support adherence and satisfaction. Therefore, more attention should be paid to the treatment experience of low-income and low-education patients. Targeted out-of-hospital care and follow-up visits should be arranged, and care for their conditions should be given to improve their treatment experience, thus improving their health.

Studies on all four domains at these three time points show that post-hospital management of CHF patients is currently lacking. In the physiological and psychological domains, patient states improve shortly after discharge, but the trend of improvement slows down and stops after a long period. In the social and therapeutical domains, patients’ states maintain stable shortly after discharge but are more likely to deteriorate after a long period. Therefore, patients’ condition in social and therapeutical domains deteriorate faster than that in physiological and psychological domains, patients require more attention after discharge. Patient adherence and family support should be maintained to improve patient psychological levels. Furthermore, close attention should be paid to the conditions of CHF patients after discharge, and the treatments should be adjusted timely for different subtypes. For instance, patients with lower incomes and education levels should be treated with more patience and care, and treatment medication should be explained in more detail to improve their treatment experience. In addition to medication, more attention should be paid to lifestyle habits, such as moderate exercise, diet, and tea drinking, to improve patients’ long-term health conditions.

This study revealed the trajectory of disease development and state transition in patients of different subtypes over time based on CHF-PROMs. The effects of different covariates, such as demographic characteristics and lifestyle, on longitudinal changes in patients’ quality of life, were compared to provide a basis for clinicians to develop personalized treatment plans according to their conditions and improve patients’ quality of life. However, there are still some limitations in this study. Some continuous variables were downgraded to categorical variables, which may cause information loss. Not all the covariates that could affect patient state transition were covered in this study, and the combined effect of multiple covariates was not studied. The research methodology should be further improved. In addition, the samples from three hospitals in Taiyuan cannot represent the whole Shanxi. The results of the study may have selection bias, and the extrapolation has certain limitations. The source data will be further optimized, the CHF sample size will be expanded, and more factors will be considered in the following studies. Furthermore, the combined effect of multiple factors and the intrinsic correlations between the four domains will be considered.

This paper demonstrates post-discharge status transition in CHF patients based on longitudinal PROMs data. For CHF patients, more attention should be paid to the social and therapeutic areas after discharge, and the post-hospital management time should be appropriately extended. This study may provide additional knowledge and increase clinical understanding regarding the early assessment and management of patients with CHF.

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

The research program received medical and ethical approval from Shanxi Medical University (No. 2018LL128). The patients/participants provided their written informed consent to participate in this study.

HY conceived the study, designed the study protocol, analyzed and interpreted the data, and drafted and wrote the manuscript. JT revised and reviewed the manuscript. JZ, JL, GH, and HY were responsible for collecting the data. HY and JZ participated in the data analysis. QH and YZ came up with the original concept for the study, oversaw the data analysis, and revised the manuscript. All authors contributed to the article and approved the submitted version.