Clinical outcomes of deceased donor kidney transplantation: Eight years of experience from a tertiary care center in India

Abstract

BACKGROUND

Deceased donor kidney transplantation (DDKT) remains underutilized in India due to organ shortage. Understanding determinants of transplant outcomes such as delayed graft function (DGF), graft survival and patient survival is essential to optimize patients and wisely allocate organs to best deserving patients.

AIM

To identify donor, recipient and peri-transplant determinants of DGF, graft and patient survival in DDKT.

METHODS

This is an ambispective cohort study for patients transplanted between January 2017 and September 2024. Primary outcome was incidence of DGF and predicting factors. Secondary outcomes were graft and patient survival. Logistic regression was used to analyse predicting factors for DGF and cox proportional hazards model was used to assess factors for graft and patient survival.

RESULTS

Among 62 DDKT, DGF occurred in 35.5% of DDKT recipients. While 6 factors affected DGF, only high recipient body mass index and high donor terminal serum creatinine were independent predictors of DGF. Over a median follow up of 32 months, 4 patients returned to maintenance haemodialysis and 11 patients died with functioning grafts with most events occurring within first year of transplant. Graft survival was 85.2% at 1 year and 68.2% at 4 years, while patient survival was 90.2% and 75.5%, respectively. Recipient age and kidney donor risk index (KDRI) were associated with graft loss. No factor affected mortality.

CONCLUSION

These findings highlight the importance of comprehensive donor and recipient assessment along with using KDRI or a new prediction tool for organ allocation in India. Larger multicentric studies from India are needed to further refine risk prediction models in DDKT.

Key Words: Deceased donor kidney transplant; Live donor kidney transplant; Delayed graft function; Graft survival; Patient survival

Core Tip: In this ambispective study we evaluated predictors of delayed graft function (DGF), graft and patient survival in deceased donor kidney transplant programme over a period of eight years in a tertiary care centre in north India. Only high recipient body mass index and high donor terminal serum creatinine were independent predictors of DGF. High recipient and donor age, kidney donor risk index and DGF were associated with graft loss.

INTRODUCTION

India accounts for nearly one-third of the world’s chronic kidney disease (CKD) population[1]. For patients progressing to maintenance haemodialysis, kidney transplantation offers a permanent cure improving quality of life. Yet, transplantation rates remain low, primarily because many patients lack compatible or willing living donors and face financial barriers. For these individuals, deceased donor kidney transplantation (DDKT) represents an alternative. However, due to shortage of organs, DDKT rates are low. It is important to recognize the value of each donated organ, to be much familiar with the modifiable factors if any that will impact upon the outcomes, to prognosticate the potential recipient likewise, and to prevent organ wastage. Choosing kidneys from deceased donors with varied health profiles and sometimes prolonged cold ischemia time (CIT), needs a clear understanding of how donor factors affect transplant outcomes such as delayed graft function (DGF), graft survival and patient survival.

Multiple studies have analysed recipient and donor factors influencing outcome after DDKT[2-4]. DGF represents an important short-term outcome which affects long term outcomes namely graft survival and patient survival. A study of 2637 recipients by Ahlmark et al[5] showed that donor and recipient’s increasing age, male sex and KDPI were associated with increased risk of DGF. USRDS registry of DDKT showed that female recipients of male donor kidneys had higher risk of graft failure[6]. A systematic review and meta-analysis by Hill et al[7] linked obesity with graft loss and mortality.

However, the applicability of these findings to the Indian context remains uncertain. Differences in donor characteristics, higher prevalence of marginal donors, variability in CIT, and unique healthcare and allocation dynamics may significantly influence transplant outcomes. Furthermore, the relative contribution and interaction of these factors in determining outcomes such as DGF, graft survival, and patient survival have not been well characterized in Indian cohorts. Addressing this gap is essential for improving risk stratification, guiding donor selection, and minimizing organ wastage in a resource-limited setting. Therefore, this study evaluates the factors influencing DGF and transplant outcomes in a real-world Indian DDKT population.

MATERIALS AND METHODS

Study design and patients

This ambispective cohort study included adult (≥ 18 years of age) patients who underwent DDKT at All India Institute of Medical Sciences, New Delhi, India between January 2017 to September 2024. The inclusion of both retrospective and prospective data in this study is intended to enhance the overall robustness and feasibility of the research. Retrospective data allow for rapid accumulation of cases, providing preliminary insights, improving statistical power, and enabling hypothesis refinement based on real-world clinical patterns. The prospective component, in contrast, ensures standardized data collection, minimizes missing variables, and allows for more rigorous assessment of outcomes and temporal relationships. Clinical data were collected retrospectively from hospital records for patients transplanted between January 2017 and June 2023 and prospectively between July 2023 and September 2024. All the adult patients who received DDKT were included in the study and none was excluded. Baseline demographic, biochemical and clinical data for both the recipient and donor were extracted. Donor details including age, sex, body mass index (BMI), terminal serum creatinine, comorbidities like hypertension, diabetes, were retrieved from hospital records. Kidney donor risk index (KDRI) and kidney donor profile index (KDPI) were calculated as per existing formulas recommended in 2025. Recipient characteristics including age, BMI, basic kidney disease, prior sensitization history of blood transfusion, pregnancy or second transplant, diabetes, peak panel reactive antibody (PRA), preoperative-type of dialysis (peritoneal vs haemodialysis). Peritransplant factors included CIT which was defined as the period from the start of perfusion with cold preservation fluid after cessation of circulation, due to arterial clamping, until the start of the first vascular anastomosis at implantation[8]. Informed consent was taken from the patients enrolled prospectively and it was waived off for the patients enrolled retrospectively. The study has been approved by Institutional Ethics Committee (IECPG-388/20.07.2023).

Outcome

The primary outcome was the incidence of DGF after DDKT and the recipient and donor factors associated with DGF. DGF was defined as the need for dialysis within the first week post-transplant[9]. Secondary outcomes included graft and patient survival and factors associated with them. Graft loss was defined as return to permanent dialysis, re-transplantation or patient death due to any cause whichever came first[10]. The patients who died with functioning grafts were also included in graft loss. Patients who were lost to follow-up were contacted telephonically to ascertain their current status regarding graft loss and death. The last follow up of patients who could not be contacted was considered based on their last visit.

Management protocol

All transplants were ABO compatible and complement-dependent cytotoxicity crossmatch negative. Donors with elevated creatinine were accepted if their baseline creatinine was documented to be normal with few donors undergoing a pre implantation kidney biopsy to rule out chronicity or cortical necrosis. The kidneys were stored in a standard cold storage solution (Graftsol). Only brain-dead donors were considered for kidney donation in our institute. All recipients received induction with rabbit-anti-thymocyte globulin and methylprednisolone which were continued for 3 days postoperatively. Maintenance of immunosuppression consisted of prednisolone, mycophenolate mofetil, and tacrolimus. All patients received prophylaxis with trimethoprim/sulfamethoxazole for 12 months, valganciclovir for 6 months and fluconazole for 1 month.

Statistical analysis

Variables were checked for normal distribution using the Shapiro-Wilk normality test. Normally distributed continuous variables were expressed as mean ± SD and skewed variables were expressed as median and interquartile range. Categorical data were reported as counts and percentages. Mann-Whitney U test was used for continuous non-normally distributed variables. χ² or Fisher’s exact test for discrete variables. Logistic regression was done to calculate odds ratio (OR) for DGF and variables having P value ≤ 0.10 were included in multivariate model. Kaplan-Meier method was used for graft survival and patient survival analysis with hazard ratio calculated using cox proportional hazards model which was tested in multivariate model for factors having P value ≤ 0.10. Given the limited number of graft loss events in addition to the missing data for KDRI, multivariable Cox regression model was not applied for KDRI in graft loss analysis. Patients with missing values for a given variable were excluded from the corresponding analysis. All the analysis has been done using Stata/MP v17.0.

RESULTS

Description of cohort

Out of 780 transplants performed in study period, 62 (7.9%) were DDKT. The baseline characteristics of recipient and donor are shown in Table 1. Recipients included 29 males (46.8%) and 33 females (53.2%), while donors included 50 males (80.6%) and 12 females (19.4%). Median age of recipients in DDKT was 42 (34-47) years. Four patients had diabetes mellitus in pre-transplant period. Underlying etiology for CKD was unknown in 45 patients with glomerulonephritis being the most common known cause in 8 patients. Other etiologies are detailed in Supplementary Table 1. Patients who could not be biopsied and imaging couldn’t confirm the diagnosis pre transplant were listed under unknown. The median donor age was 45 (26-58) years. All except one donor who died from spontaneous intracerebral haemorrhage had traumatic brain death. The median terminal serum creatinine was 1.1 (0.8-1.7) and median CIT was 3.5 (2.7-4.9) hours. A total of 62 patients were included (45 retrospective, 17 prospective), with greater missing data in the retrospective cohort: HLA was missing in 7 patients (6 vs 1), PRA in 10 (10 vs 0), KDRI in 14 (13 vs 1), and CIT in 4 (3 vs 1). Patients with missing values for a given variable were excluded from the corresponding analysis. The retrospective cohort had a longer follow-up duration compared to the prospective cohort [median 37 (28-85) vs 15 (10-19) months].

Table 1 Recipient and donor baseline characteristics in non-delayed graft function and delayed graft function groups.

Variables	Total (n = 62)	No-DGF (n = 40)	DGF (n = 22)	P value
Recipient characteristics
Age	42 (34-47)	41 (34-44)	43 (35-49)	0.286
Male	29 (46.8)	15 (37.5)	14 (63.6)	0.048
BMI	21.8 (19.2-23.4)	20.6 (18.7-22.7)	22.8 (21.5-26.5)	0.013
Diabetes mellitus	4 (6.5)	2 (5.0)	2 (9.1)	0.610
Dialysis vintage (months)	81 (63-96)	83 (71-102)	76 (50-84)	0.035
Peritoneal dialysis	7 (11.3)	3 (7.5)	4 (18.1)	0.233
PRA class I > 20 (n = 52)	9/52 (17.3)	6/34 (17.7)	3/18 (16.7)	1.000
PRA class II > 20 (n = 52)	14/52 (26.9)	10/34 (29.4)	4/18 (22.2)	0.746
Any sensitisation event ever	54 (87.1)	33/40 (82.5)	21/22 (95.5)	0.240
Pregnancy (n = 33)	23/33 (69.7)	16/25 (64.0)	7/8 (87.5)	0.382
Blood transfusion	44 (71.0)	26 (65.0)	18 (81.8)	0.243
Second transplant	8 (12.9)	3 (7.5)	5 (22.7)	0.119
Donor characteristics
Age	45 (26-58)	45 (25-53)	52 (32-64)	0.077
Male	50 (80.6)	35 (87.5)	15 (68.2)	0.065
Terminal creatinine	1.1 (0.8-1.7)	1.0 (0.8-1.3)	1.4 (0.8-2.2)	0.014
BMI	24.5 (23.3-26.3)	24.5 (23.0-26.4)	24.8 (24.2-26.3)	0.504
Hypertension	12 (19.4)	7 (17.5)	5 (22.7)	0.618
Diabetes mellitus	11 (17.7)	6 (15.0)	5 (22.7)	0.446
Expanded criteria donor	14 (22.6)	7 (17.5)	7 (31.8)	0.197
HLA mismatch (n = 55)				0.137
0, 1, 2	0	0	0
3	3/55 (5.5)	2 (5.4)	1 (5.6)
4	9/55 (16.4)	3 (8.1)	6 (33.3)
5	21/55 (38.2)	16 (43.2)	5 (27.8)
6	22/55 (40.0)	16 (43.2)	6 (33.3)
Donor KDRI Rao (n = 48)	1.2 (1.0-1.6)	1.1 (1.0-1.4)	1.4 (1.1-1.7)	0.098
KDRI scaled (2024) (n = 48)	0.8 (0.7-1.1)	0.8 (0.7-1.0)	1.0 (0.8-1.2)	0.098
KDPI 2025 (%) (n = 48)	32 (15-61)	24 (14-47)	48 (23-67)	0.145
Cold ischemia time (hours) (n = 58)	3.5 (2.7-4.9)	3.4 (2.6-4.1)	4.7 (3.0-6.0)	0.054
Donor-recipient sex combinations				0.011
Male to male	24/62 (38.7)	15/24 (62.5)	9/24 (37.5)
Male to female	26/62 (41.9)	20/26 (76.9)	6/26 (23.1)
Female to male	5/62 (8.1)	0	5/5 (100)
Female to female	7/62 (11.3)	5/7 (71.4)	2/7 (28.6)

Data is presented as n (%) or median (interquartile range). BMI: Body mass index; PRA: Panel reactive antibody; HLA: Human leukocyte antigen; KDPI: Kidney donor profile index; KDRI: Kidney donor risk index; DGF: Delayed graft function.

Incidence and predictors of DGF

DGF was seen in 22 (35.5%) patients. There was significantly higher risk of DGF in male recipients compared to females (P = 0.048), patients with higher BMI (P = 0.013) and lower dialysis vintage (P = 0.035). Raised terminal donor creatinine also showed significantly higher risk of DGF (P = 0.014). No significant difference in occurrence of DGF was seen with recipient age, pretransplant diabetes, type of dialysis, pretransplant PRA, any sensitisation event, donor BMI, hypertension, diabetes or HLA mismatch. There was a trend toward a lower incidence of DGF among recipients of male donors than female donors (P = 0.065). Donor-recipient sex combinations showed a statistically significant effect on DGF (P = 0.011). The incidence of DGF was highest in female donor to male recipient transplants (100%), whereas the lowest rate was observed in male donor to female recipient pairs (23.1%). A trend towards statistical significance was also seen with higher KDRI and KDPI and longer CIT (Table 1).

We employed three different multivariate models adjusting for different variables (Table 2). Model 1 adjusted for core baseline clinical covariates (recipient BMI, recipient gender, dialysis vintage, donor age, donor sex, donor terminal creatinine). Model 2 incorporated KDRI instead of individual components. Model 3 adjusted for CIT separately as CIT represents perioperative management rather than donor/recipient core baseline biology. Overall, these findings suggest that increased BMI of recipients and kidneys from donors with higher terminal creatinine are at substantially increased risk of DGF independently, whereas other donor and recipient characteristics were not independently predictive after adjustment. Recipient male gender shows an increased odds of DGF in all three multivariate models, but the statistical significance varies across models.

Table 2 Multivariate logistic regression analysis of predictors of delayed graft function.

Variable	Model 1		Model 2		Model 3
	OR (95%CI)	P value	OR (95%CI)	P value	OR (95%CI)	P value
Recipient BMI	1.28 (1.07-1.53)	0.007	1.16 (0.97-1.40)	0.099	1.22 (1.01-1.46)	0.036
Male recipient	4.31 (1.00-18.52)	0.050	2.81 (0.63-12.50)	0.174	2.85 (0.56-14.57)	0.207
Dialysis vintage (months)	1.00 (1.00-1.00)	0.833	1.00 (1.00-1.00)	0.853	1.00 (1.00-1.00)	0.826
Male donor	0.28 (0.04-1.99)	0.205	0.24 (0.04-1.60)	0.141	0.32 (0.05-2.32)	0.264
Donor age	1.02 (0.98-1.06)	0.326			1.04 (0.99-1.09)	0.137
Donor terminal creatinine	2.66 (1.25-5.66)	0.011			2.29 (1.09-4.82)	0.029
KDRI			1.60 (0.31-8.25)	0.574
CIT (hours)					1.00 (1.00-1.01)	0.133

Model 1 calculated for 62 patients, model 2 calculated for 48 patients as kidney donor risk index available for 48 patients and model 3 calculated for 58 patients as cold ischemia time available for 58 patients. BMI: Body mass index; KDRI: Kidney donor risk index; CIT: Cold ischemia time; OR: Odds ratio.

Graft loss and mortality

Over a median follow up of 32 months (17-71 months), 4 patients returned to maintenance haemodialysis and 11 patients died with functioning grafts. Most adverse events occurred in the first-year post-transplant, where 2/4 (50.0%) patients experienced graft loss and 6/11 (54.5%) patients died with functioning grafts. Graft survival was 85.2% (95%CI: 73.6-92.1) at 1 year, 81.3% (95%CI: 68.6-89.2) at 2 years, 68.2% (95%CI: 50.9-80.5) at 4 years. Patient survival was 90.2% (95%CI: 79.5-95.5) at 1 year, 86.2% (95%CI: 74.1-92.9) at 2 years, 75.5% (95%CI: 58.2-86.5) at 4 years.

DGF was associated with a substantially higher risk of graft failure, with patients experiencing DGF having a 3.07-fold greater risk compared with those without DGF (95%CI: 1.09-8.62; P = 0.034), as is shown in Figure 1A. DGF also showed a trend towards increasing mortality risk [hazard ratio (HR) 3.20, 95%CI: 0.94-10.95; P = 0.063] as is shown in Figure 1B. The trends were lost after adjusting for recipient and donor characteristics with a HR for graft loss (Table 3) being 2.50 (95%CI: 0.85-7.34; P = 0.095) and for mortality 2.40 (95%CI: 0.65-8.87; P = 0.191).

Figure 1 Effect of delayed graft function on graft survival and patient survival. A: Graft survival; B: Patient survival. DGF: Delayed graft function; HR: Hazard ratio.

Table 3 Multivariate Cox regression analysis of predictors of graft loss.

Variable	Model 1		Model 2
	HR (95%CI)	P value	HR (95%CI)	P value
Recipient age	1.06 (1.00-1.12)	0.024	1.07 (1.00-1.13)	0.022
Donor age	1.03 (1.00-1.07)	0.051	1.02 (1.00-1.06)	0.137
DGF			2.50 (0.85-7.34)	0.095

Model 1 (n = 62) - recipient and donor age. Model 2 (n = 62) - Model 1 + delayed graft function - this model incorporates baseline clinical covariates along with delayed graft function which represents immediate post operative graft function. Collinearity was checked using variance inflation factor (VIF) and variables with VIF > 5.0 (kidney donor profile index) were excluded. DGF: Delayed graft function; KDRI: Kidney donor risk index; HR: Hazard ratio.

On univariable analysis (Supplementary Table 2), higher donor risk indices were significantly associated with graft loss. KDRI demonstrated a strong association with both graft loss and mortality (Supplementary Tables 2 and 3). But this association did not persist for mortality (Supplementary Table 4, HR: 2.72, 95%CI: 0.71-10.46; P = 0.144). Recipient age was also significantly associated with increasing graft loss (Table 3, HR: 1.06, 95%CI: 1.00-1.12; P = 0.024) but not mortality.

DISCUSSION

In this ambispective cohort study of DDKT from a high-volume tertiary care centre in India, DGF occurred in more than one-third of recipients and was independently associated with BMI and donor terminal creatinine. In addition, higher donor risk indices (KDRI/KDPI) and increasing recipient age, donor age were associated with decreased graft survival. Importantly, DGF was associated with an increased risk of graft loss, although this association attenuated after adjustment for donor and recipient factors. These findings highlight key donor and recipient determinants influencing early and long-term outcomes following DDKT in the Indian setting.

The incidence of DGF in our cohort (35.5%) is similar to that reported in many western registries despite the differences in donor characteristics, allocation practices, and recipient factors across transplant programme[5,11]. In India, deceased donor transplantation often occurs under significant logistical constraints, including limited donor availability and variable organ retrieval networks[12]. Consequently, transplant centres may accept organs from donors with less optimal characteristics, which may lead to DGF despite the relatively short CIT (as median 3.5 hours in our cohort). CIT has been shown to be associated with increased risk of DGF[5,13-15] but it did not influence DGF in our study. The absence of a statistically significant association between CIT and DGF in our cohort should be interpreted in the context of the remarkably short median CIT. Most of donor were from in house trauma centre which resulted in short and stable CIT.

Recipient BMI emerged as an independent predictor of DGF. Obesity has been consistently associated with inferior transplant outcomes, including DGF and graft loss[16]. Several mechanisms have been proposed to explain this association. Obesity is linked to wound infection, dehiscence, increased chances of hyperglycemia. Also, obesity is characterised by chronic systemic inflammation, endothelial dysfunction, and increased oxidative stress, all of which may exacerbate ischemia-reperfusion injury in the transplanted kidney[16], which predisposes the graft to early dysfunction. Previous studies have reported similar associations. A large cohort study by Molnar et al[17] demonstrated a significant relationship between higher BMI and DGF (adjusted HR: 1.34). A point to be noted is that the median BMI in that study was 26.8 kg/m², however our median BMI was 21.8 kg/m² indicating that association holds even within non-obese cohort. BMI may act as a continuous risk factor reflecting underlying metabolic and inflammatory susceptibility to ischemia-reperfusion injury. The attenuation of this effect after KDRI adjustment indicates interaction with donor quality. While obesity is linked to increased chances of DGF as shown in other studies, its effects on long term outcomes is controversial. When compared to non-obese recipients, some reports described an increased risk of graft failure and mortality for obese recipients whilst others have found no significant differences[7,18,19]. Contemporary cohorts from Korea similarly report an association between BMI and DGF without a corresponding effect on graft or patient survival[20].

Similarly, elevated donor creatinine likely reflects underlying donor kidney injury, increasing vulnerability to reperfusion stress[21]. We observed that terminal creatinine and donor acute kidney injury do increase the risk of DGF but doesn’t have impact on graft survival and mortality. Similar findings have been reported in other studies also[22-24].

Interestingly, male recipients showed a higher risk of DGF, although this association was not consistently statistically significant across multivariate models. Donor-recipient sex combinations also revealed that incidence of DGF was highest in female donor to male recipient transplants, whereas the lowest rate was observed in male donor to female recipient pairs. These findings suggest that sex mismatch, particularly the female-to-male combination, may predispose to early graft dysfunction. In female-donor-to-male-recipient transplants, reduced nephron mass from the donor kidney is coupled with a higher metabolic demand and a testosterone-driven pro-inflammatory milieu in the recipient. This convergence likely amplifies susceptibility to ischemic injury, thereby predisposing to DGF[25]. Experimental murine models demonstrating testosterone-mediated exacerbation of IRI provide mechanistic support for this observation, and our clinical data extend these findings by illustrating how such biological differences manifest as markedly increased DGF risk in human transplant settings[26]. Deceased donor renal transplants (n = 128493) from UNOS data (1997-2011) and by Budhiraja et al[27] showed that male recipient gender was highly associated with DGF [OR: 1.39 (1.33-1.46)] independent of donor gender. However, the above donor-recipient sex combination data must be interpreted cautiously given the extremely small sample size especially female to male donation (only 5 such instances were there).

According to large population databases from Finland and United States[5], DGF, KDPI, diabetes, recipient age and sex, peak PRA affected graft loss. Consistent with existing literature, DGF was associated with an increased risk of graft loss in our study also. DGF reflects early ischemia-reperfusion injury, which can trigger inflammatory cascades, endothelial dysfunction, and activation of innate immunity, ultimately predisposing to acute rejection and chronic allograft injury[9]. Although the association between DGF and graft failure lost statistical significance after multivariable adjustment in our study, the direction and magnitude of effect remained consistent, suggesting that DGF remains an important marker of graft vulnerability. KDRI and KDPI, were associated with graft survival in our cohort. Our findings confirm the applicability of these risk indices in predicting transplant outcomes even within an Indian population.

Recipient age was also independently associated with graft loss. Older recipients may have higher comorbidity burden, reduced physiological reserve, and greater susceptibility to complications such as infections or cardiovascular events, which may adversely influence long-term outcomes after transplantation. This observation underscores the importance of careful recipient selection and individualized risk assessment in deceased donor transplantation. Indian study by Gopalakrishnan et al[4] also reported recipient age as an independent predictor for graft and patient survival at 3 years.

Our study provides important insights into DDKT outcomes in India complementing existing reports from other Indian centres. But unlike prior Indian studies that described only outcomes post DDKT, our study specifically examined impact of donor and recipient clinical characteristics along with peri-transplant factors (CIT, DGF) on graft and patient survival. Application of multivariable analysis provides more robust results compared to earlier descriptive studies which used only univariate analysis. The study also reflects real-world clinical practice, with all eligible patients included and minimal selection bias. The study reflects outcomes in current era of immunosuppression and donor management making these findings more relevant to present-day clinical practice. Lastly, the ambispective design allowed capture of both historical and prospective data, enabling a comprehensive evaluation of transplant outcomes over time.

KDRI and KDPI demonstrated an association with graft survival in our cohort, suggesting that donor quality indices developed in Western populations may retain some prognostic relevance in the Indian deceased donor setting. However, these findings should be interpreted cautiously. The KDRI was originally derived and validated using large United States transplant registry data and incorporates donor characteristics, demographic variables, and weighting systems specific to that population. Certain components of the original model, including body metrics, and causes of donor death, may not fully reflect the demographic and clinical characteristics of Indian deceased donors.

In the Indian context, deceased donors are generally younger, predominantly trauma-related, and differ substantially from Western donor populations in terms of comorbidity burden, anthropometric characteristics, and healthcare infrastructure. Therefore, the predictive performance and calibration of KDRI may differ when applied to Indian recipients.

Larger multicentric Indian studies are required to evaluate the external validity of KDRI and potentially develop a region-specific donor risk prediction model better suited to the Indian transplant population.

Missing KDRI and CIT data were mainly confined to the retrospective cohort and were attributable to incomplete archival donor and perioperative records, particularly during the earlier transplant period before standardized data collection practices were established.

A major limitation of the present study is the relatively small sample size and limited number of graft loss events, which restrict the statistical power of multivariable Cox regression analyses. Although clinically relevant covariates were selected carefully and collinearity was assessed prior to model construction, the possibility of model overfitting cannot be excluded. Therefore, the identified associations should be interpreted cautiously and considered exploratory rather than definitive independent predictors. Larger multicentric studies with higher event rates are needed to validate these findings and better define the determinants of graft outcomes following deceased donor kidney transplantation.

While this study represents one of the larger DDKT series from India, single centre study limits generalizability and several associations might not have reached statistical significance despite showing clinical trends due to small sample size. Retrospective data are subject to selection bias, incomplete or inconsistent documentation, and potential misclassification of variables. Additionally, differences in data collection methods between retrospective and prospective phases may introduce heterogeneity. There is also a risk of information bias and confounding, particularly if key variables were not uniformly recorded in the retrospective cohort. As most patients were studied retrospectively, certain donor variables such as haemodynamic parameters, vasopressor requirements, biopsy findings, vascular anatomy, warm ischemia times were not consistently available for analysis. Prior studies suggest that dopamine infusion in donor before organ retrieval may reduce the risk for DGF but such data is unavailable for our cohort. These factors are known to significantly influence renal perfusion and the degree of ischemic injury prior to organ retrieval. Their absence limits our ability to fully interpret the relationship between donor terminal creatinine and the development of DGF. The observational design precludes causal inference. Finally, the follow-up duration, although adequate for assessing early outcomes, may be insufficient to fully evaluate long-term graft survival. The relatively low number of mortality events and limited follow-up duration reduce the statistical power to detect such associations for mortality. These findings suggest that while donor and recipient factors are important determinants of early graft outcomes, their influence on patient survival is likely multifactorial and requires larger studies with longer follow-up for definitive assessment.

CONCLUSION

In conclusion, DGF remains common following DDKT in our setting and is influenced by both donor and recipient characteristics. Higher recipient BMI, male sex and elevated donor terminal creatinine increase the risk of DGF, while higher donor risk indices and increasing recipient age are associated with poorer graft survival. The observed association between KDRI and graft loss in our cohort suggests its potential applicability in risk stratification within the Indian transplant setting. However, given differences in donor demographics and clinical profiles compared to western populations, a modified allocation model tailored to the Indian donor pool may be more appropriate. The development of such a tool would require large, multicentric datasets and prospective validation before clinical implementation. Our analysis could serve as a hypothesis generating rather than hypothesis confirming.