Chronic kidney disease progression modeled using a continuous-time multistate Markov process revealed increasing likelihood of progression to advanced stages over time, with diabetes, hypertension, and cardiovascular disease as primary factors associated with disease progression.
Key Findings
Results
Transition probabilities indicated an increasing likelihood of progression to more advanced CKD stages over time, accompanied by decreasing probabilities of remaining in the same stage.
Study followed 194 patients with CKD across 1506 clinic visits at Jimma Medical Center between February 2019 and February 2024
A retrospective cohort design was used with a continuous-time multistate Markov framework
The model estimated transition intensities, transition probabilities, mean sojourn times, and next-state probabilities across CKD stages
Results
Mean sojourn times differed across CKD stages, with stage 3 having the longest expected duration.
Mean sojourn time in stage 1 was 4.33 months
Mean sojourn time in stage 2 was 4.44 months
Mean sojourn time in stage 3 was 5.79 months
Mean sojourn time in stage 4 was 4.57 months
Results
Patients with hypertension were significantly more likely to transition from CKD stage 4 to stage 5 (kidney failure) compared to those without hypertension.
Hazard ratio for transition from stage 4 to stage 5 among hypertensive patients was 2.62
95% confidence interval: 1.79–3.84
This was described as a statistically significant association
Results
Diabetes, hypertension, and cardiovascular disease were identified as the primary clinical factors associated with CKD progression.
These three comorbidities were assessed as covariates within the multistate Markov framework
Both clinical and demographic characteristics were evaluated for their effects on transitions between CKD stages
Hypertension showed a quantified significant effect (HR = 2.62) specifically for the stage 4 to stage 5 transition
Results
Mean sojourn times and next-state probabilities provided complementary information to transition intensities by describing expected stage duration and likelihood of subsequent transitions.
Next-state probabilities described the likelihood of transitioning to each subsequent CKD stage from a current stage
Mean sojourn times described the expected duration spent in each CKD stage
Together, these metrics offered a more complete picture of disease dynamics than transition intensities alone
The authors concluded these findings 'may support stage-specific management and intervention strategies'
Background
Conventional survival models were deemed inadequate for capturing CKD progression, motivating the use of a multistate Markov framework.
CKD involves multiple intermediate stages between initial diagnosis and kidney failure
The continuous-time multistate Markov model was selected to handle transitions among multiple states simultaneously
The framework allowed estimation of stage-specific transition dynamics rather than only time to a single endpoint
What This Means
This research followed 194 patients with chronic kidney disease (CKD) over five years at Jimma Medical Center in Ethiopia, recording 1,506 clinic visits to understand how the disease progresses through its stages. Instead of using traditional methods that only look at a single outcome like death or kidney failure, the researchers used a sophisticated statistical approach called a multistate Markov model, which can track movement between multiple disease stages over time. This allowed them to estimate how long patients typically stay in each stage and how likely they are to move to a worse stage.
The study found that, over time, patients became increasingly likely to progress to more advanced stages of CKD. Patients spent an average of roughly 4 to 6 months in each stage, with stage 3 having the longest average duration at about 5.79 months. Having hypertension (high blood pressure) was a particularly strong risk factor: hypertensive patients were 2.62 times more likely to progress from stage 4 to stage 5 (kidney failure) compared to those without hypertension. Diabetes and cardiovascular disease were also identified as important drivers of disease progression.
This research suggests that modeling CKD as a multi-stage process reveals important details about how quickly the disease advances at each stage and which patient characteristics speed up that progression. These findings could help clinicians identify patients at highest risk of rapid deterioration and tailor monitoring and treatment strategies to specific disease stages, particularly targeting management of hypertension, diabetes, and cardiovascular disease to slow CKD progression.
Futasa Begna T, Tereda A, Abdisa Fufa J, Diriba T, Debusho L. (2026). Modeling chronic kidney disease progression using a continuous-time multistate Markov model: Transition dynamics and covariate effects.. The Journal of international medical research. https://doi.org/10.1177/03000605261477526