Pragmatic cardiovascular-kidney-metabolic burden categories and 5-year all-cause mortality in Vietnamese outpatients: A retrospective cohort study.

BACKGROUND

The American Heart Association introduced cardiovascular-kidney-metabolic (CKM) health as an integrated framework for metabolic, kidney, and cardiovascular risk. However, full AHA CKM staging requires variables that are often unavailable in retrospective outpatient datasets, and outcome data from Southeast Asia remain limited.

OBJECTIVE

To evaluate whether pragmatic CKM burden categories derived from routinely available baseline diagnoses identify Vietnamese outpatients at higher risk of 5-year all-cause mortality.

METHODS

We performed a retrospective cohort analysis of 480 adult outpatients recruited from 01 January 2016-31 December 2016 in Ho Chi Minh City, Vietnam. The de-identified dataset for this secondary analysis was accessed on 31 March 2024. Participants were classified into four mutually exclusive pragmatic CKM burden categories: Category A, no documented metabolic-risk diagnosis, chronic kidney disease (CKD), or coronary artery disease (CAD); Category B, documented metabolic-risk diagnosis only; Category C, CKD or CAD, but not both; and Category D, concomitant CKD and CAD. These categories are AHA-informed but are not official AHA CKM stages. The primary endpoint was 5-year all-cause mortality, with administrative censoring at 5 years. Logistic regression was the primary inferential model; Kaplan-Meier, log-rank, and Cox models were used as complementary time-to-event analyses. To support privacy-preserving data sharing, adjusted models used prespecified age groups (<40, 40-60, and >60 years) and sex rather than exact individual ages.

RESULTS

Most participants were aged 40-60 years (331/480, 69.0%) and 204 (42.5%) participants were men. The category distribution was as follows: Category A, 24 (5.0%), Category B, 252 (52.5%), Category C, 180 (37.5%), Category D, 24 (5.0%). Over 2243.1 person-years of follow-up within the 5-year analysis horizon, 64 deaths occurred (13.3%). Five-year mortality was 0/24 (0.0%) in Category A, 33/252 (13.1%) in Category B, 24/180 (13.3%) in Category C, 7/24 (29.2%) in Category D. Kaplan-Meier curves differed across the four categories (log-rank p = 0.036). Compared with Categories A + B, Category D had higher unadjusted odds of 5-year mortality (OR 3.03, 95% CI 1.17-7.86; p = 0.022) and remained elevated after age-group and sex adjustment (OR 2.87, 95% CI 1.04-7.94; p = 0.042). Category C was not associated with higher adjusted 5-year mortality.

CONCLUSIONS

In this Vietnamese outpatient cohort, pragmatic CKM burden categories identified a subgroup with combined CKD and CAD that had the highest absolute 5-year mortality. The age-group- and sex-adjusted estimate remained elevated for this small subgroup, although confidence intervals were wide. These categories should not be interpreted as official AHA CKM stages, and prospective validation with complete CKM phenotyping is needed.