For many complex surgical procedures, there is an association between a low volume of procedures and an increased risk of death. Liver transplantation (LT) is regarded as one of the most complex surgical procedures, thus LT outcomes may be influenced by the LT team experience. In the United States (US), transplant centers performing ≤20 LT cases per year showed a higher mortality rate than the higher LT volume centers [1]. In contrast, it was reported that institutional case volume was not a significant predictor of post-transplant survival in the model for the end-stage liver disease (MELD) era [2]. In the US studies to examine whether individual surgeon volume is associated with LT outcomes, there was no significant relationship between LT volume and LT outcomes at the transplant center level [3,4]. These data suggest that the quality of LT teams performing deceased donor liver transplantation (DDLT) may be equalized in most of the US LT centers.

On the other hand, a cohort study using Korea-nationwide data revealed better in-hospital mortality (IHM) and long-term survival after DDLT in the high-volume LT centers compared to low-volume centers [5]. Another Korea-nationwide retrospective cohort study also revealed that high-volume LT centers also had better outcomes after living donor liver transplantation (LDLT) and pediatric LT compared with low-volume centers [6,7]. In 2020, 1,616 cases of LT, 46 DDLT, and 51 LDLT cases, were performed in 56 centers in Korea [8]. Considering that the majority of LT centers in Korea are classified as low-volume centers, it is necessary to evaluate the effect of institutional LT cases volume on post-transplant outcomes through a meta-analysis.

A total of 1,616 LT cases were performed in 56 centers in the year 2020. DDLT and LDLT were performed in 46 and 51 centers, respectively. Five centers performed DDLT only, and their LT volume was one or two cases. Ten centers performed LDLT only, with LT volume of one or two cases in eight centers, but 8 and 28 cases each were reported at two centers. Forty-one centers performed both DDLT and LDLT (Fig. 1) [8].

Figure 1. Distribution of Korean transplantation centers performing deceased donor liver transplantations (DDLT) and living donor liver transplantations (LDLT) in 2020. LT, liver transplantation.

In the year 2020, 31 (55.4%) centers performed 1 to 10 LTs; 13 (23.2%) centers reported 11 to 25 LTs; 7 (12.5%) centers performed 26 to 50 LTs; one (1.8%) center conducted 55 LTs; three centers had LT volume of 118, 139 and 151 each; and one (1.8%) center reported 511 LTs. The latter five major centers performed 974 LTs, which accounted for 60.3% of the LT volume nationwide in the year 2020. The 511 LTs performed at a single center constituted 31.6% of the nationwide LT volume. These major five LT centers performed 179 cases of DDLT, which accounted for 45.3% of the nationwide DDLT volume; and also performed 795 LDLTs, which constituted 65.1% of the nationwide LDLT volume (p<0.001). These findings suggest that LDLT was predominantly performed at these high-volume LT centers [8].

A nationwide retrospective cohort study was performed to investigate the relationship between institutional case volume and post-transplant outcomes after DDLT. The data was extracted from the Korean National Healthcare Insurance Service (NHIS) database. A total of 2,648 adult DDLTs were performed at 54 centers from January 2007 to December 2016. Centers were divided into high-, medium-, and low-volume centers according to the average annual number of DDLTs: <10, 10–30, and >30, respectively. The IHM rates in high-, medium-, and low-volume centers were 10.3%, 14.3%, and 17.1%, respectively (Fig. 2). Multivariable logistic regression analysis revealed that low-volume centers (adjusted relative risk [RR]=1.953; 95% confidence interval [CI]=1.461–2.611; p<0.001) and medium-volume centers (adjusted RR=1.480; 95% CI=1.098–1.994; p=0.010) had a significantly higher IHM rate compared to high-volume centers. Long-term mortality rates were also higher in low-volume centers (p=0.007) (Fig. 3). This study concluded that centers with higher volume showed better IHM and long-term survival after DDLT compared to centers with lower volume [5].

Figure 2. Relationship between the average annual number of deceased donor liver transplantation (DDLT) and in-hospital mortality at 54 centers, from 2007 to 2016. High-volume center: >30 liver transplantations (LTs)/year, Medium-volume center: 10–30 LTs/year, Low-volume center: <10 LTs/year.

Figure 3. Kaplan-Meier survival curve after adult deceased donor liver transplantation (DDLT) according to the center case volume.

A nationwide retrospective cohort study using the Korean NHIS database aimed to evaluate the relationship between institutional case volume and clinical outcomes after adult LDLT. Between January 2007 and December 2016, 7,073 adult LDLTs were performed at 50 centers in Korea. Centers were categorized as high-, medium-, and low-volume centers according to the average annual number of LTs: >50, 10 to 50, and <10, respectively. The IHM rates in the high-, medium-, and low-volume centers were 2.8%, 4.1%, and 6.7%, respectively (Fig. 4). After adjustment, the IHM rate was significantly higher in the low-volume centers (adjusted RR=2.287; 95% CI=1.471–3.557; p<0.001) and medium-volume centers (adjusted RR=1.676; 95% CI=1.089–2.578; p=0.019) compared with high-volume centers (Fig. 5). The high-volume centers showed that long-term survival for up to 9 years was better, and intensive care unit and hospital length of stay were shorter. This study concluded that centers with higher cases volume (>50 LTs/year) had better outcomes after adult LDLT, including IHM and long-term mortality compared with centers with lower case volumes (≤50 LTs/year) [6].

Figure 4. Relationship between the average annual volume and in-hospital mortality after adult living donor liver transplantation. High-volume center: >50 liver transplantations (LTs)/year, Medium-volume center: 10–50 LTs/year, Low-volume center: <10 LTs/year.

Figure 5. Kaplan-Meier survival curve after adult living donor liver transplantation according to the center case volume.

A nationwide retrospective cohort study using the Korean NHIS database was conducted to evaluate whether institutional cases volume affects clinical outcomes after pediatric LT. Between January 2007 and December 2016, 521 pediatric LTs were performed at 22 centers in Korea. Centers were categorized as high-, medium-, and low-volume centers according to the average annual number of LTs: >10, 1 to 10, and <1, respectively. The IHM rates in the high-, medium-, and low-volume centers were 5.8%, 12.5%, and 32.1%, respectively (Fig. 6). After adjustment, the IHM rate was significantly higher in the low-volume centers (adjusted RR=9.693; 95% CI=4.636–20.268; p<0.001) and medium-volume centers (adjusted RR=3.393; 95% CI=1.980–5.813; p<0.001) compared to high-volume centers. Long-term survival for up to 9 years was better in high-volume centers (Fig. 7). This study concluded that centers with higher cases volume (>10 pediatric LTs/year) had better outcomes after pediatric LT, including IHM and long-term mortality, compared to centers with lower case volumes (≤10 LTs/year) [7].

Figure 6. Average annual case volume and in-hospital mortality after pediatric liver transplantation (LT). High-volume center: >10 LTs/year, Medium-volume center: 1–10 LTs/year, Low-volume center: <1 LTs/year.

Figure 7. Kaplan-Meier survival curve after pediatric liver transplantation (LT) according to the center case volume.

The relationship between institutional LT case volume and clinical outcomes after liver re-transplantation was assessed through analysis of the patients who underwent liver re-transplantation between 2007 and 2016. They were selected from the Korean NHIS database. LT centers were categorized to either high-volume centers (≥64 LTs/year) or low-volume centers (<64 LTs/year) according to the annual LT case volume. A total of 258 liver re-transplantations were performed during the study period: 175 liver re-transplantations were performed in three high-volume centers and 83 were performed in 21 low-volume centers. The IHM rates after liver re-transplantation in high and low-volume centers were 25% and 36%, respectively (p=0.069; Fig. 8). Adjusted IHM rate was not different between low and high-volume centers. Adjusted 1-year mortality was significantly higher in low-volume centers (RR=2.14; 95% CI=1.05–4.37, p=0.037) compared to high-volume centers. Long-term survival for up to 9 years was also higher in the high-volume centers (p=0.005; Fig. 9). Other risk factors of IHM and 1-year mortality included female gender and higher Elixhauser comorbidity index. This study concluded that centers with higher case volume (≥64 LTs/year) showed lower IHM and overall mortality after liver re-transplantation compared to low-volume centers [9].

Figure 8. Relation between the institutional cases volume and in-hospital mortality after liver re-transplantation. High-volume center: ≥64 LTs/year, Low-volume center: <64 LTs/year. LT, liver transplantation.

Figure 9. Kaplan-Meier survival curve after liver re-transplantation according to the center case volume. LT, liver transplantation.

Long-term mortality rates of the above-mentioned four cohort studies for DDLT, adult LDLT, pediatric LT, and liver re-transplantation were analyzed [5-7,9]. Meta-analysis of these data showed that high-volume LT centers had lower mortality rates compared with low-volume centers (RR=0.710; 95% CI=0.533–0.911; Fig. 10).

Figure 10. Fig. 10 . Forest plot of four Korea-nationwide cohort studies on LT. DDLT, deceased donor liver transplantation; LDLT, living donor liver transplantation; LT, liver transplantation; RR, relative risk; CI, confidence interval.

Four Korea-nationwide cohort studies on DDLT, adult LDLT, pediatric LT, and liver re-transplantation using the Korean NHIS database revealed that high-volume LT center had lower mortality rates compared with low-volume centers [5-7,9]. These Korean studies disagree with those of the high-volume US cohort studies using the registry data from the Organ Procurement Transplantation Network [2-4].

The impact of institutional cases volume on post-transplant outcomes may differ depending on the complexity of the surgical procedure. The cases volume effect on postoperative clinical outcome have been reported in complex surgical procedures [10], while institutional case volume did not influence clinical outcomes in relatively simple surgical procedures [11]. Therefore, complex surgical procedures have are preferably performed in high-volume centers with regards to patient outcomes. The present study analysis results suggest that higher institutional case volume is associated with improved short- and long-term survival after LT.

The results of the present meta-analysis may be used as evidence supporting centralization of LT. Patients requiring complex and high-risk surgical procedures may anticipate better outcomes when they receive care in designated centers with sufficient experience. Our previous study revealed that the number of LT centers in Korea was 56 in the year 2020 [8]. Forty-one transplant centers performed both DDLT and LDLT, five centers performed only DDLT, and 10 centers conducted only LDLT. Five major centers performed 974 cases, which accounted for 60.3% of nationwide LT volume in 2020. These centers performed 45.3% of all the nationwide DDLT volume and 65.1% of the nationwide LDLT volume in 2020, indicating that LDLT was highly concentrated at these high-volume LT centers [8].

There is no clear cutoff between the high and low-volume centers based on patient outcome after LT. Considering that the LDLT surgical techniques have been gradually standardized in Korea and those of DDLT are well established already worldwide [12-15], it is expected that with the accumulation of institutional experience in small-volume LT centers, the difference in post-transplant outcomes in the near past data would be diluted in the near future. The old cohort studies in the US indicated the presence of inter-center differences [1], but similar studies after the year 2000 have reported the absence of such differences, especially after the adoption of the MELD score system [2,3]. Liver re-transplantation is a demanding procedure, much more difficult than primary LT. A US cohort study revealed that center volume is an imprecise surrogate measure for 1-year outcomes after liver re-transplantation [4].

The completeness of the NIHS database in coverage of the whole Korean population is a unique strength of the four cohort studies included in the present meta-analysis. The Korea healthcare system is a single-payer system and insures more than 97% of residents in Korea [12] with equal benefits, regardless of the insurance premiums that differ depending on income. The remaining 3% of the population with the lowest income is supported by the Medical Aid program. The bulk of the incurred medical expenses are reimbursed by the NHIS and the details of the claim is stored in the NHIS database. Therefore, the four cohort studies are free from selection bias, or incomplete/missing data regarding the outcomes. Due to the completeness of the Korean NHIS database, the results of the present study may reflect the real-world outcomes after LT.

In conclusion, the results of the present meta-analysis revealed lower mortality rates in high-volume LT centers compared with low-volume centers regarding all type of LT, including DDLT, adult LDLT, pediatric LT, and liver re-transplantation. Considering that the surgical techniques for LDLT have been gradually standardized and those for DDLT are established already, it is expected that with the accumulation of institutional experience in small-volume LT centers, the difference in post-transplant outcomes experienced in the near past data would be compensated in the near future.

There was no funding related to this study.

All authors have no conflicts of interest to declare.

Conceptualization: YKC. Data curation: All. Methodology: All. Visualization: All. Writing - original draft: All. Writing - review & editing: All.