Background: Mortality in end-stage kidney disease (ESKD) remains high, especially among the elderly with a higher burden of comorbidity and frailty. In this group, dialysis may not offer better survival compared to conservative management. Frailty defined by clinical frailty scale (CFS) and comorbidity by Charlson Comorbidity Index (CCI) are known independent predictors of mortality. Our aim was to compare which one had higher impact on early mortality in urgent-start hemodialysis patients.

Methods: We conducted a retrospective cohort study of patients aged 65 years and over, who started hemodialysis as their first renal replacement therapy (RRT) between January 2014 and December 2020. CFS and mCCI, at time of HD start, were used to evaluate, respectively, frailty and comorbid disease burden. The primary outcome was death in the first 6 months of RRT.

Results: During the study period 166 patients were included. The median age, at time of hemodialysis start, was 75 years ± 6.3 years. The mortality at 6 months was 19% (n=31). For both scales, the analysis of ROC curve, stablished the optimal cut-off to predict the event death at first 6 months as ≥ 5points. The performance of CFS was superior to the mCCI, in fact, the area under the curve was significantly higher in CFS (0.739) versus the mCCI (0.620). A CFS≥5 had a sensitivity/specificity of 94%/44% in prediction the primary outcome. On the other hand, a mCCI≥5 predicts the same outcome with a sensitivity/specificity of 26%/88%. The diagnostic odds ratio for CFS≥5 was 11.6, compared to only 2.7 for mCCI≥5. Lastly, we constructed a model in which both scores interacted (as categorical variables), which after multivariable adjustment showed that mCCI/CFS <5/≥5 and ≥5/≥5 were independent predictors of mortality (HR=7.06; P=0.017; HR=10.708; P=0.002; respectively). Interestingly, no events were observed in the mCCI≥5/CFS<5 group.

Conclusions: In this cohort of urgent-start incident HD patients, frailty defined by CFS was a stronger predictor of mortality than comorbidity defined by CCI.

Keywords: frailty, comorbidity, elderly, dialysis, chronic kidney disease

Despite the considerable proportion of health care resources being used, morbidity and mortality remain high in end-stage kidney disease (ESKD) patients especially during the first months on dialysis therapy.^1–3 Comorbidity and frailty, both frequent in elderly patients with chronic kidney disease (CKD), are associated with a higher risk of death even irrespective of age and dialysis need.^4–7

Frailty is a measure of increased vulnerability to stressful situations caused by reduced functional biological reserves.⁸ This condition raises the risk of progression of disease, disability, hospitalization, and death.⁹ There are many tools to assess the degree of frailty, clinical frailty scale (CFS) is a simple yet global frailty assessment, proved as an independent predictor of mortality in dialysis patients.^10–12 The Charlson Comorbidity Index (CCI), a scale used to assess and quantify comorbidity burden, has also proved to have prognostic value in a variety of conditions as CKD.^5,13,14 Frailty and comorbidity are often presumed to be close concepts however they don’t always coexist, even sometimes traducing opposite information about the same patient. We aimed to access the predictive value of frailty and comorbidity to predict early death in urgent start hemodialysis patients.

We conducted a retrospective cohort study of patients aged 65 years and over, referred to the Nephrology Department of Centro Hospitalar Universitário do Porto (CHUPorto), who underwent urgent start hemodialysis as their first renal replacement therapy (RRT), between January 2014 and December 2020. CHUPorto is a tertiary care hospital, which serves a diverse population of 500,000 inhabitants in the Northern region of Portugal. Urgent start hemodialysis was defined as any first treatment started for an emergency condition or not appropriate to delay for >24 h, even if a permanent dialysis access was in place.

Data were collected from medical records and included (at dialysis initiation): sex, age, associated comorbid conditions, such as diabetes, dyslipidemia, hypertension, smoking status, history of malignancy, and cardiovascular disease. Variables related to renal care included vascular access (graft/fistula versus catheter) and the timing of nephrologist care referral prior to dialysis. Glomerular filtration rate (GFR) was estimated using the CKD Epidemiology 2021 creatinine equation15; all serum creatinine measurements were performed in the same laboratory calibrated using a calibrator for automated systems (Roche Diagnostics).

A modified version of the Charlson comorbidity index (mCCI), by excluding subject’s age and presence of kidney disease, was used to evaluate comorbidity. Frailty was defined by the Clinical Frailty Scale that evaluates specific domains including comorbidity, function, and cognition to generate a frailty score ranging from 1 (very fit) to 9 (terminally ill).

The study outcome was all-cause mortality within first 6 months following dialysis therapy initiation. Vital status was checked until August 30, 2021.

Continuous data were described using mean (standard deviation) or median (interquartile range) and categorical data were expressed as numbers (frequencies). Categorical were compared using Pearson χ 2 test or Fisher's exact test, as appropriate. Continuous variables were compared with Student t-test or Mann–Whitney U test), as appropriate. Receiver operator characteristic (ROC) curves and area under the curve (AUC) were calculated for CFS and mCCI scores, as well as sensitivity, specificity, positive and negative diagnostic odds ratios. Graft survival curves were visualized using Kaplan–Meier method, with comparison groups being done by log-rank test. Association between frailty and comorbidity index (and their interaction) and early death was performed by multivariable Cox regression model.

A two-sided P-value of <0.05 was considered statistically significant. Statistical calculations were performed using Stata/MP, version 15.1 (Stata Corp, College Station, TX).

Study population

During the study period, 166 patients with 65 years old or more met the criteria of urgent-start HD. Patient demographics are shown in Table 1. The median age was 75.5 years, 58% were men. Common comorbidities included hypertension (97%), dyslipidemia (93%), congestive heart failure (72%), diabetes (54%) and peripheral vascular disease (51%).

Comorbidity and frailty

Figure 1 shows the distribution of CFS and mCCI scores. The median CFS was 5 (inter quartile range 4-6) The median mCCI was 4 (inter quartile range 3-6).

Figure 1 Distribution of the mCCI and CFS in the cohort. mCCI, modified Charlson comorbidity index; CFS, clinical frailty scale.

Considering 6-month mortality prediction, the performance of CFS was superior to the mCCI (p=0.031; Figure 2). In fact, the area under the curve (AUC) was considerably higher in CFS (0.739) versus the mCCI (0.620). For both scores, the analysis of receiver operating characteristics curve, established the optimal cut-off to be ≥ 5points. A CFS≥5 had a sensitivity/specificity of 94%/44% in prediction the primary outcome, respectively. On the other hand, a mCCI≥5 predicted the same outcome with a sensitivity/specificity of 26%/88%. The diagnostic odds ratio for CFS≥5 was 11.6, compared to only 2.7 for mCCI≥5. The group of frail patients (CSF ≥5) were older (p=0.016), started dialysis with a higher eGFR (p=0.005) and had more comorbidities (p<0.001) (Table 1).

Figure 2 Area under the Receiver operating characteristic curves by mCCI and CFS. mCCI, modified Charlson comorbidity index; CFS, clinical frailty scale.

Survival analysis

Nineteen percent of the patients died during the follow up period. Table 1 shows the clinical and demographic data, accordingly vital status. In this cohort, age was not significantly different (p=0.540), while later referral to nephrology (p=0.005), lower serum albumin (p=0.049) and catheter as vascular access (p=0.037) were more common in those who died. Overall survival rate at 3/6-months was 89/81%, respectively.

Patients’ survival curves by CFS (≥5 vs <5 at 3-/6-months, respectively: 83%/72% vs 98%97%; P<0.001) and mCCI (≥5 vs <5 at 3-/6-months, respectively: 85%/74% vs 92%/88%; P=0.020) groups are depicted in Figure 3 A & B, respectively. Survival curves stratified by both scores are shown in Figure 4, both groups with CFS≥5 had worse survival, independently from the mCCI score.

Figure 3 Kaplan Meier survival curves by CFS (A) and mCCI (B). mCCI, modified Charlson comorbidity index; CFS, clinical frailty scale.

Figure 4 Kaplan-Meier survival curves by groups defined by the interaction of mCCI and CFS. mCCI , modified Charlson comorbidity index; CFS, clinical frailty scale.

In a multivariable adjusted cox model (Table 2), CFS was an independent predictor of 6-month mortality both as a categorical (CSF≥5) (HR=3.64; P=0.004) and continuous variable (HR=1.95; P<0.01); while mCCI was not. When groups defined by the interaction of both scores were analyzed (CFS/mCCI <5/<5 as reference), in group 1 (mCCI≥5/CFS<5) no events were observed, while the remaining groups (mCCI>5/CFS≥5, mCCI≥5/CFS≥5) were independent predictors of mortality.¹⁵

In this study, we found that frailty was a better predictor of early mortality in urgent-start hemodialysis patients.

Whereas the best definition and classification of frailty and comorbidity remain a matter of discussion, it is well recognized that they entail a higher risk of unfavorable outcomes such as hospitalization and death.^9,16 Despite the higher risk of death by virtue of having CKD, frailty and comorbidity were associated with increased mortality. In our study, this association persisted after adjustment for age, sex, body mass index, serum albumin, vascular access and estimated glomerular filtration rate at dialysis start (Tables 3 & 4).

Frailty and comorbidity were frequent in our population, in congruence with similar cohorts.^17–23 Almost two thirds of the patients were frail, probably because CKD is a risk factor to frailty but also because these patients had a high burden of comorbidities. Some symptoms associated with comorbidities, such as weakness and slowness, can prompt physicians to misdiagnose a frailty phenotype, on the other hand it is also likely that these comorbidities can weaken patient’s biological reserves.²⁴ We confirmed that mCCI and CFS are independent predictors of mortality, CFS as a categorical and continuous variable but mCCI only as a categorical variable. The increased risk of death with continuous increase in the CFS score emphasizes the potential prognostic significance of frailty. Our findings are comparable with those of Rockwood et al and Alfaadhel et al, both observed a similar increased risk of death.^11,25 This risk increase was not achieved by mCCI.

In this cohort of incident hemodialysis patients aged 65 years and over, almost a fifth died within the first 6 months of follow-up. These results are slightly superior to those found in previous studies.^26–29 We can speculate about the reasons for this difference, namely our patients were older and demanded urgent start hemodialysis which are two known conditions for worse outcomes. Nonetheless, age at dialysis start, did not modified the association between frailty, comorbidity, and death. This finding was reported in some other studies and shows that additional factors influence patient’s outcomes regardless of age.^7,30

We intended to access the ability of mCCI and CFS to predict mortality but also to evaluate the interaction between these scores. Both scores had a DOR higher than 1, but CFS≥5 DOR was 4 times higher than mCCI≥5. Also, Kaplan-Meier analysis by CFS and mCCI shows earlier and more expressive separation of the curves in the CFS compared to mCCI. These data support that CFS is a better score to predict early mortality versus the mCCI. Frailty ability to predict mortality is evident even in the cases where mCCI traduces low comorbidity burden as showed by the model in which both scores interacted as categorical variables. Is also important to point out that the group with CFS and mCCI <5 had almost 100% survival rate at 3 and 6 months proving the ability of these scores to predict not only death but also good outcomes.

The exact reason why frailty predicts mortality better then comorbidity is unknown. Frailty evaluates the level of physiologic reserves of one individual that may be decreased by their comorbidities, sometimes despite the high burden of diseases the functional status is preserved. Our study pointed out that the repercussion of the diseases is a better determinant of early death than the diseases alone.

This work has also some limitations. First, it is a single-center study with a limited number of patients. Despite the effort to adjust the results, the observational nature of the study might allow the existence of unmeasured and unknow confounders that influence the relation between frailty, comorbidity, and death. CFS is an impression-based frailty score, its subjective nature might promote the existence of some bias. It can also be argued that instead of comparing frailty and comorbidity we compared two different scores that might not properly evaluate frailty and comorbidity. Nonetheless, we choose two scores that are already extensively validated in hemodialysis patients.^{6,10,11,31–33} Finally, as one time acquisition of the scores also enables the misclassification of the patients, longitudinal evaluation of frailty and comorbidity could improve the accuracy of data collection and limit the misclassification.

In conclusion, we identified that frailty and comorbidity defined by CFS and mCCI are associated with an increased early mortality risk on incident dialysis patients. We also proved frailty to be a more accurate predictor of early mortality versus comorbidity especially in the first three months. Further studies should be held to validate these conclusions which supports that decisions related with RRT should be based on frailty evaluation instead of chronological age and or/comorbidity alone.

None.

The authors report no conflicts of interest in this work.

None.

1. Saran R, Robinson B, Abbott KC, et al. US Renal Data System 2019 Annual Data Report: Epidemiology of Kidney Disease in the United States. Am J Kidney Dis. 2020;75(1):A6–A7.
2. Heaf J, Heiro M, Petersons A, et al. First ‑ year mortality in incident dialysis patients: results of the Peridialysis study. BMC Nephrol. 2022;1–12.
3. Davison SN. End-of-life care preferences and needs: Perceptions of patients with chronic kidney disease. Clin J Am Soc Nephrol. 2010;5(2):195–204.
4. McAdams Demarco MA, Tan J, Salter ML, et al. Frailty and cognitive function in incident hemodialysis patients. Clin J Am Soc Nephrol. 2015;10(12):2181–2189.
5. Manoch Rattanasompattikul, Usama Feroze, Miklos Z Molnar, et al. Charlon comorbidity score is a strong predictor of mortality in hemodialysis patients. Int Urol Nephrol. 2012;44(6):1813–1823.
6. Pugh J, Aggett J, Goodland A, et al. Frailty and comorbidity are independent predictors of outcome in patients referred for pre-dialysis education. Clin Kidney J. 2016;9(2):324–329.
7. Bao Y, Dalrymple L, Chertow GM, et al. Frailty, dialysis initiation, and mortality in end-stage renal disease. Arch Intern Med. 2012;172(14):1071–1077.
8. Schulte Kemna L, Künzig M, Dallmeier D, et al. Frailty in renal diseases. Z Gerontol Geriatr. 2021;54(7):708–716.
9. Vermeiren S, Vella Azzopardi R, Beckwée D, et al. Frailty and the Prediction of Negative Health Outcomes: A Meta-Analysis. J Am Med Dir Assoc. 2016;17(12):1163.e1–1163.e17.
10. Imamura K, Yamamoto S, Suzuki Y, et al. Comparison of the association between six different frailty scales and clinical events in patients on hemodialysis. Nephrol Dial Transplant. 2022.
11. Rockwood K, Song X, MacKnight C, et al. A global clinical measure of fitness and frailty in elderly people. CMAJ. 2005;173(5):489–495.
12. Kumarasinghe AP, Chakera A, Chan K, et al. Incorporating the Clinical Frailty Scale into routine outpatient nephrology practice: an observational study of feasibility and associations. Intern Med J. 2021;51(8):1269–1277.
13. Wu PH, Lin YT, Lee TC, et al. Predicting Mortality of Incident Dialysis Patients in Taiwan - A Longitudinal Population-Based Study. PLoS One. 2013;8(4):1–7.
14. Ng YY, Hung YN, Wu SC, et al. Progression in comorbidity before hemodialysis initiation is a valuable predictor of survival in incident patients. Nephrol Dial Transplant. 2013;28(4):1005–1012.
15. Inker LA, Eneanya ND, Coresh J, et al. New Creatinine- and Cystatin C–Based Equations to Estimate GFR without Race. N Engl J Med. 2021;385(19):1737–1749.
16. Hemmelgarn BR, Manns BJ, Quan H, et al. Adapting the Charlson comorbidity index for use in patients with ESRD. Am J Kidney Dis. 2003;42(1 SUPPL. 2):125–132.
17. Li Y, Zhang D, Ma Q, et al. The impact of frailty on prognosis in elderly hemodialysis patients: A prospective cohort study. Clin Interv Aging. 2021;16:1659–1667.
18. Takeuchi H, Uchida HA, Kakio Y, et al. The prevalence of frailty and its associated factors in Japanese hemodialysis patients. Aging Dis. 2018;9(2):192–207.
19. Delgado C, Shieh S, Grimes B, et al. Association of Self-Reported Frailty with Falls and Fractures among Patients New to Dialysis. Am J Nephrol. 2015;42(2):134–140.
20. Johansen KL, Dalrymple LS, Delgado C, et al. Association between body composition and frailty among prevalent hemodialysis patients: A US renal data system special study. J Am Soc Nephrol. 2014;25(2):381–389.
21. Garcia Canton C, Rodenas A, Lopez Aperador C, et al. Frailty in hemodialysis and prediction of poor short-term outcome: mortality, hospitalization, and visits to hospital emergency services. Ren Fail. 2019;41(1):567–575.
22. Chowdhury R, Peel NM, Krosch M, et al. Frailty and chronic kidney disease: A systematic review. Arch Gerontol Geriatr. 2017;68:135–142.
23. Collard RM, Boter H, Schoevers RA, et al. Prevalence of frailty in community-dwelling older persons: A systematic review. J Am Geriatr Soc. 2012;60(8):1487–1492.
24. Musso CG, Jauregui JR, Macías Núñez JF. Frailty phenotype and chronic kidney disease: a review of the literature. Int Urol Nephrol. 2015;47(11):1801–1807.
25. Alfaadhel TA, Soroka SD, Kiberd BA, et al. Frailty and mortality in dialysis: Evaluation of a clinical frailty scale. Clin J Am Soc Nephrol. 2015;10(5):832–840.
26. Song YH, Cai GY, Xiao YF, et al. Risk factors for mortality in elderly haemodialysis patients: A systematic review and meta-analysis. BMC Nephrol. 2020;21(1):1–10.
27. Hazara AM, Bhandari S. Early Mortality Rates After Commencement of Maintenance Hemodialysis: A Systematic Review and Meta-Analysis. Ther Apher Dial. 2020;24(3):275–284.
28. Santos J, Fonseca I, Malheiro J, et al. End-stage renal disease versus death in a Portuguese cohort of elderly patients: An approach using competing event analysis. J Investig Med. 2017;65(7):1041–1048.
29. Arif FM, Sumida K, Molnar MZ, et al. Early Mortality Associated with Inpatient versus Outpatient Hemodialysis Initiation in a Large Cohort of US Veterans with Incident End-Stage Renal Disease. Nephron. 2017;137(1):15–22.
30. McAdams Demarco MA, Law A, Salter ML, et al. Frailty as a novel predictor of mortality and hospitalization in individuals of all ages undergoing hemodialysis. J Am Geriatr Soc. 2013;61(6):896–901.
31. Pinto MP, Martins S, Mesquita E, et al. European portuguese version of the clinical frailty scale: Translation, cultural adaptation and validation study. Acta Med Port. 2021;34(13):749–760.
32. Lee J, Jung J, Lee J, et al. Recalibration and validation of the Charlson Comorbidity Index in acute kidney injury patients underwent continuous renal replacement therapy. Kidney Res Clin Pract. 2022;41(3):332–341.
33. Van Manen JG, Korevaar JC, Dekker FW, et al. How to adjust for comorbidity in survival studies in ESRD patients: A comparison of different indices. Am J Kidney Dis. 2002;40(1):82–89.