Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2021 Jan;21(1):229-240.
doi: 10.1111/ajt.16142. Epub 2020 Aug 4.

Validation of the Model for End-stage Liver Disease sodium (MELD-Na) score in the Eurotransplant region

Ben F J Goudsmit  1   2   3 , Hein Putter  4 , Maarten E Tushuizen  3 , Jan de Boer  2 , Serge Vogelaar  2 , I P J Alwayn  1 , Bart van Hoek  3 , Andries E Braat  1
Affiliations

Validation of the Model for End-stage Liver Disease sodium (MELD-Na) score in the Eurotransplant region

Ben F J Goudsmit et al. Am J Transplant. 2021 Jan.

Abstract

The MELD score is used in the Eurotransplant (ET) region to allocate liver grafts. Hyponatremia in cirrhotic patients is an important predictor of death but is not incorporated in MELD. This study investigated the performance of the MELD-Na score for the ET region. All adult patients with chronic liver disease on the ET liver transplantation waiting list (WL) allocated through lab MELD scores were included. The MELD-corrected effect of serum sodium (Na) concentration at listing on the 90-day WL mortality was calculated using Cox regression. The MELD-Na performance was assessed with c-indices, calibration per decile and Brier scores. The reclassification from MELD to MELD-Na score was calculated to estimate the impact of MELD-Na-based allocation in the ET region. For the 5223 included patients, the risk of 90-day WL death was 2.9 times higher for hyponatremic patients. The MELD-Na had a significantly higher c-index of 0.847 (SE 0.007) and more accurate 90-day mortality prediction compared to MELD (Brier score of 0.059 vs 0.061). It was estimated that using MELD-Na would reduce WL mortality by 4.9%. The MELD-Na score yielded improved prediction of 90-day WL mortality in the ET region and using MELD-Na for liver allocation will very likely reduce WL mortality.

Keywords: clinical research / practice; liver disease; liver transplantation / hepatology; liver transplantation: auxiliary; mathematical model; organ allocation; organ procurement and allocation; recipient selection.

PubMed Disclaimer

Conflict of interest statement

The authors of this manuscript have no conflicts of interest to disclose as described by the American Journal of Transplantation.

Figures

Figure 1
Figure 1
The flowchart of in‐ and exclusion for this study
Figure 2
Figure 2
Violin plots with embedded box plots of the median serum sodium (Na) levels at listing, for the most frequent causes of liver disease. The dotted line represents the median Na of 137 mmol/L for the whole cohort. For the significant differences between Na levels, P values for pairwise comparisons are shown
Figure 3
Figure 3
Cumulative incidence plots for 90‐day WL outcomes, with competing risks of death, transplantation and removal due to clinical condition or censoring for NSE or HU status during waiting. Hyponatriemic patients show increased rates of mortality (27%) and transplantation (33%) compared to normonatriemic patients (respectively 8% and 18%) patients. For an explanation of the NSE and HU status, see Jochmans et al
Figure 4
Figure 4
Generalized additive Cox model with spline showing the effect of serum sodium at listing on 90‐day mortality, corrected for the MELD score
Figure 5
Figure 5
The concordance statistics (c‐indices) for 90‐day mortality of MELD and MELD‐Na between 2007 and 2018
Figure 6
Figure 6
Calibration plot of the MELD‐Na model showing the predicted and observed risks of death per decile (10%) of the patient population. The diagonal line represents a perfect calibration
Figure 7
Figure 7
A, Heatmap of the gained MELD‐Na points for each combination of MELD and serum sodium level at listing. For example, MELD 24 patients with a Na level of 125 mmol/L gain an average of 6 points, resulting in a MELD‐Na score of 30. (Also available online as interactive plot.) B, Heatmap of the difference in predicted death probability between MELD and MELD‐Na. For each MELD and Na combination the gained predicted risk from MELD‐Na is shown, which represents which patients will be prioritized most (red area, highest predicted risk difference) and which patients will lose the most priority with MELD‐Na‐based allocation (blue area, lowest predicted risk difference). (301394302also available online as interactive plothttps://plot.ly/~Liver_Research/3/ and https://plot.ly/~Liver_Research/5/)
Figure 8
Figure 8
Reclassification from MELD (y‐axis with percentage of patients with that score) to MELD‐Na (x‐axis with percentage of patients with that score). The diagonal shows which patients remain in the same stratum, that is, not reclassified, and which patients are reclassified to a higher MELD‐Na score (percentages in the tiles). A lighter color indicates a higher difference between MELD and MELD‐Na scores. The histogram on the left shows for each MELD score the average gain in MELD‐Na points
See this image and copyright information in PMC

Comment in

  • MELD calibration.
    D'Amico G, Maruzzelli L. D'Amico G, et al. Am J Transplant. 2021 Jan;21(1):438-439. doi: 10.1111/ajt.16255. Epub 2020 Aug 27. Am J Transplant. 2021. PMID: 32786172 No abstract available.
  • Invited response to "MELD calibration".
    Goudsmit BFJ, Putter H, Tushuizen ME, de Boer J, Vogelaar S, Alwayn IPJ, van Hoek B, Braat AE. Goudsmit BFJ, et al. Am J Transplant. 2021 Jan;21(1):440-441. doi: 10.1111/ajt.16289. Epub 2020 Sep 19. Am J Transplant. 2021. PMID: 32893951 Free PMC article. No abstract available.

References

    1. Eurotransplant . Annual Report 2017. 2018. www.eurotransplant.org. Accessed March 3, 2019.
    1. Ng M, Fleming T, Robinson M, et al. Global, regional, and national prevalence of overweight and obesity in children and adults during 1980–2013: A systematic analysis for the Global Burden of Disease Study 2013. Lancet. 2014;384(9945):766‐781. - PMC - PubMed
    1. Asrani SK, Devarbhavi H, Eaton J, Kamath PS. Burden of liver diseases in the world. J Hepatol. 2019;70(1):151‐171. - PubMed
    1. Pimpin L, Cortez‐Pinto H, Negro F, et al. Burden of liver disease in Europe: Epidemiology and analysis of risk factors to identify prevention policies. J Hepatol. 2018;69(3):718‐735. - PubMed
    1. Estes C, Razavi H, Loomba R, Younossi Z, Sanyal AJ. Modeling the epidemic of nonalcoholic fatty liver disease demonstrates an exponential increase in burden of disease. Hepatology. 2018;67(1):123‐133. - PMC - PubMed