Trajectories of Hospitalization Cost Among Patients of End-Stage Lung Cancer: A Retrospective Study in China

Zhong Li, Shan Jiang, Ruibo He, Yihan Dong, Zijin Pan, Chengzhong Xu, Fangfang Lu, Pei Zhang, Liang ZhangInternational Journal of Environmental Research and Public Health 2018 · DOI

TL;DR

This study uses Latent Class Analysis (LCA) to identify distinct trajectories of end-of-life hospitalization costs for lung cancer patients in China, revealing three cost patterns: low and increasing, high and decreasing, and moderate and decreasing. Understanding these trajectories and their associated determinants (education, hospitalization frequency, place of death) can inform targeted palliative care interventions and improve resource allocation.

What Problem Does This Paper Address?

End-of-life (EOL) healthcare costs for cancer patients represent a significant global concern, with a substantial portion of lifetime healthcare spending occurring in the final months of life. Aggressive medical treatment during this period drives these costs, placing a strain on healthcare resources, especially in developing countries where access to palliative care is often limited. While previous studies have examined EOL costs, there's a need to understand the heterogeneity in cost trajectories and the factors influencing these patterns to enable more effective and timely palliative care interventions. This paper specifically addresses this by investigating the diverse ways hospitalization costs evolve for end-stage lung cancer patients in China.

Methods

The researchers employed a two-pronged statistical approach:

  • Latent Class Analysis (LCA): This is the core methodology used to identify unobserved heterogeneity within the patient population. LCA groups individuals into distinct "classes" based on their observed patterns of behavior – in this case, their hospitalization costs over time. The study used a censored normal model with the traj procedure, suitable for data with frequent zero values and cost increases. The Bayesian Information Criterion (BIC) was used to determine the optimal number of classes.
  • Multi-nominal Logit Model: Once the classes were identified, this model was used to explore the determinants of class membership. It helps understand which demographic and clinical factors are associated with a patient belonging to a particular cost trajectory.

Datasets

The study utilized data from the National Population Death Registration and Management System and the Yichang Health Management Centre in urban Yichang, China. This integrated dataset included information on:

  • Demographics of cancer decedents (place of death, type of cancer, date of death).
  • Hospital information systems.
  • Health insurance databases.
  • Population information databases.

The final study sample consisted of 134 end-stage lung cancer patients who had survived for at least six months prior to death.

Key Insights and Ideas

  • Cost Trajectories Aren't Uniform: The most significant insight is that end-of-life hospitalization costs for lung cancer patients are not monolithic. LCA revealed distinct patterns, moving beyond simple averages to capture the dynamic evolution of spending.
  • Acceptance vs. Aggression: The differing trajectories suggest a spectrum of patient and family engagement with the dying process. Class 1's increasing costs might reflect a continued pursuit of aggressive treatment, while Classes 2 and 3's declining costs could indicate a shift towards palliative care or acceptance of impending death, though the high costs in these classes suggest this transition might be late.
  • Education Matters for EOL Care: A higher level of education was associated with being in the higher-cost trajectory (Class 3). This could imply better understanding of treatment options, greater access to information, or potentially more assertive engagement with the healthcare system.
  • Place of Death as a Proxy: Dying at home was associated with lower costs, suggesting that home-based palliative care might be more cost-effective than hospital-based care, aligning with global trends and recommendations.
  • Hospitalization Frequency is a Strong Predictor: Increased hospitalization frequency directly correlated with higher cost trajectories, highlighting the impact of frequent acute care episodes on overall EOL spending.

Key Results

  • Three Distinct Cost Trajectories:
    • Class 1 (Persistently Low and Increasing): Represented 25.4% of patients. These patients had minimal costs that increased in the last two months of life, often indicating early discharge and subsequent readmission.
  • Class 2 (High and Decreasing): Comprised 50.0% of patients. These individuals incurred high costs that decreased as death approached.
  • Class 3 (Moderate and Decreasing): Accounted for 24.6% of patients. This group also experienced high costs that declined with impending death, but with a higher average cost than Class 2. Notably, 97% of Class 3 patients died in a medical institution.
  • Determinants of Class Membership:
    • Education: Patients with at least a senior school diploma had a significantly higher probability of being in Class 3 (higher cost) compared to Classes 1 and 2.
  • Place of Death: Dying at home was associated with a decreased probability of being in the higher-cost Classes 2 and 3, compared to Class 1.
  • Hospitalization Frequency: Each additional unit of hospitalization increased the probability of being in Classes 2 and 3 compared to Class 1.
  • Model Fit: The multi-nominal logistic regression model achieved an R² of 39.69%, indicating adequate model fitness.

Limitations and Future Work

The authors acknowledge several limitations:

  • Sample Size: The study involved a relatively small sample size (n=134), which may limit the generalizability of findings.
  • Marginally Significant p-values: Some p-values in the determinant analysis were marginally significant, suggesting caution in interpretation.
  • Data Limitations: The study did not capture preference-related, family, or community-level variables, which could also influence EOL costs.

Future research directions suggested include:

  • Validation: Validating the identified model in clinical settings with larger, more diverse datasets.
  • Risk/Protective Factors: Identifying additional risk or protective factors that can inform clinical decision-making.
  • Determinants of Early Discharge/Aggressive Care: Further investigating the specific factors that lead to early discharge or the decision for aggressive end-of-life care.

Summary updated

Abstract

This study was conducted to investigate the trajectory of hospitalization costs, and to assess the determinants related to the membership of the identified trajectories, with the view of recommending future research directions. A retrospective study was performed in urban Yichang, China, where a total of 134 end-stage lung cancer patients were selected. The latent class analysis (LCA) model was used to investigate the heterogeneity in the trajectory of hospitalization cost amongst the different groups that were identified. A multi-nominal logit model was applied to explore the attributes of different classes. Three classes were defined as follows: Class 1 represented the trajectory with minimal cost, which had increased over the last two months. Classes 2 and 3 consisted of patients that incurred high costs, which had declined with the impending death of the patient. Patients in class 3 had a higher average cost than those in Class 2. The level of education, hospitalization, and place of death, were the attributes of membership to the different classes. LCA was useful in quantifying heterogeneity amongst the patients. The results showed the attributes were embedded in hospitalization cost trajectories. These findings are applicable to early identification and intervention in palliative care. Future studies should focus on the validation of the proposed model in clinical settings, as well as to identify the determinants of early discharge or aggressive care.
BibTeX
@article{li2018trajectories,
  author = {Zhong Li and Shan Jiang and Ruibo He and Yihan Dong and Zijin Pan and Chengzhong Xu and Fangfang Lu and Pei Zhang and Liang Zhang},
  title = {{Trajectories of Hospitalization Cost Among Patients of End-Stage Lung Cancer: A Retrospective Study in China}},
  journal = {International Journal of Environmental Research and Public Health},
  year = {2018},
  doi = {10.3390/ijerph15122877},
}
Figure