Nowadays, since there is an escalating demand for new services, which unfortunately coexists with constraints of new resources (Goeree, R., & Diaby, V. 2013) [4], the financial sustainability of the modern health care system is challenging. It is well known that the intensive care unit (ICU) is one of the hospital settings that absorb most of the resources [5]. Recently, two groups of authors tried to determine if the involvement of cost considerations in the ICU clinical decision-making could be useful [6] & [7]. Even if their position in this regard was opposite, they both concluded that results of health economics analysis are indeed valuable and, if correctly applied, they can bring benefits to the patients and the community, improving the quality of care.

Among the most frequent conditions in the ICU patients, severe acute kidney injury (AKI) is one of the most problematic; it increases morbidity, mortality and length of hospital stay (Klarenbach, S. et al., 2009) [2], and its incidence has been rising in the last decades.  The most recent guidelines state that CRRT is the gold standard treatment for patients with severe AKI and hemodynamic instability, acute brain injury or increased intracranial pressure (G. Kidney Disease: 2012) [8], However, this treatment is hindered by its high cost, which is due to the devices and the qualified professional staff needed [7]. 

Given these considerations, the objective of the present study was to evaluate the costs of continuous renal replacement therapy (CRRT) and its effectiveness in terms of mortality in patients with severe AKI.

This study was designed as retrospective and multi-center. 

The goal was to calculate the costs of CRRT in critical patients and assess its effectiveness by confronting the ICU and the post-ICU outcomes with the mortality predicted by the SOFA Score.

Patient’s data have been extracted from the “GiViTI-Italian group for the evaluation of treatments in Intensive Care” database. All adult patients (age≥18) treated with CRRT in the participating ICUs (189) in the years 2011-2014, were enrolled. To increment the homogeneity of the cohort, a SOFA Score ranging from 8 to 12 was employed as an additional inclusion criterion. In the end, the study population comprised of 2599 patients. 

To obtain an estimate of the total cost of CRRT, the regional tariff (DRG) 3995.5 (€ 255, 70), applied in 2015 in the “Santa Maria della Misericordia” University hospital in Perugia, was employed. Subsequently, the DRG was multiplied by the median length of treatment (4 days) and the number of patients, to have an approximation of the resources needed to treat the study population.

To evaluate the effectiveness of CRRT, we tested if the actual outcomes were in line with the mortality predicted by the SOFA Score. Employing the findings from the study performed by Vincent et al., (1998) [3], we assumed that a SOFA Score between 8 and 12 predicted a risk of mortality ranging from 22, 5% to 48, 5%. The Fischer’s exact-test was used to compare the mortality rate during the ICU stay and the one 30 days after the discharge. A result of P<0.05 was considered significant. The statistical analysis was performed using GraphPad Prism software, ver­sion 7.02 for Windows [9].

Population characteristics

In the study cohort (Tab. 1), 63, 2% of patients were males and 36, 7% females, and their median age was 71. In most cases the BMI index classified them as normal (37, 7%) or overweight (33, 4%), whereas obese patients were 24, 9% and underweight ones only 3, 5%.

The reason for their admission to the ICU was almost always receiving intensive care; only 10, 2% were checked in to be monitored or to be weaned from artificial ventilation, and 1, 3% because they needed specific devices or treatments. 

Comorbidities affected 93,3% of the population and, in terms of frequency, the top three were blood hypertension, type 2 Diabetes Mellitus and moderate or severe acute kidney injury.

At the moment of admission, organ failures were present in 99,6% of cases.

The AKIN criteria were employed to diagnose and classify renal impairment. According to them, severe AKI was present in 1612 patients, moderate AKI in 492 and mild AKI in 333. 

During the ICU stay, complications occurred in 70,8% of cases and organ failures in 51,8%.

Endpoints

The estimated total cost of CRRT amounted to € 2 658 257, 20. 

In the ICU, the mortality rate was 44, 4%. An additional 10,5% of the population died in the hospital 30 days after ICU discharge; and the overall mortality rate was 54,9%. The Fischer’s exact-test proved that the difference in the outcomes was significant with a p<0, 0001 (Tab 2; Fig.1). 

Analyzing this sum in regard to the patients mortality, we found that € 1 339 868 was spent for patients who survived 30 days after the ICU discharge, € 1 179 288, 40 for patients who died in the ICUs; and € 139 160,80 for patients who survived in the ICU, but then died in the hospital within the following 30 days (Fig.2). 

The results of the present study indicate that the total cost of CRRT was estimated to be € 2.658.257,20. To obtain this estimate, we used the Perugia University hospital DRG, as we assumed it would be similar enough to the ones applied all over Italy, as these tools are developed with the aim to include all the real cost. 

Additionally, we found that CRRT was effective in the ICU, where the mortality (44,4%) was in line with the prediction based on the SOFA Score (22-48,5%). However, this benefit vanished after 30 days of ICU discharge, because mortality increased by 10,5% and became higher than what was predicted.

The study was designed with the assumption that, if effective, CRRT should decrease the SOFA score and, with it, the mortality risk. In reality, this tool was initially conceived to describe organ impairment in septic patients, and our findings may be biased by an improper use of the score. However, SOFA has been found to relate well with mortality by several studies (Minne, L. et al., 2008), [10] and we felt this was the most appropriate score to consider, as it has been studied both in the ICU and in the hospital wards [11 & 12]. Moreover, since AKI is usually precipitated and coexistent with other organ failures, it has been proven more predictive than other scores in critical patients with AKI [13]. So, we believe that, in these circumstances, SOFA was the most reliable score to predict both ICU and hospital short-term mortality. Ultimately, the present study is not the first one that used the SOFA score to evaluate the effectiveness of CRRT. In 2006, A. M. Rizal et al.,. already used the reduction of SOFA Score as unit of health interest and effective measure [14]. However, unlike our study, theirs was a single-center randomized clinical trial that aimed to compare high volume and standard dose hemofiltration.

Conversely, to try to shed light on the reasons behind our results, it could be useful to consider what some recently published studies concluded. Fagugli et al., (2015) [15] reported that, despite the boost of AKI incidence in the University hospital “Santa Maria della Misericordia” in Perugia, the resources allocated to the Nephrology department did not increase accordingly. They observed that in the years 2007-2012 an escalating number of patients with severe AKI were treated in medical or surgical departments other than Nephrology, and that the opportunity to dispense them with appropriate treatment were limited. Moreover, Abraham et al., (2012) [16] concluded that in English hospitals where a Nephrology team was not present, 30 day mortality after AKI was significantly higher. 

Additionally, the UK National Confidential Enquiry into Patient Outcomes and Death (NCEPOD) Adding Insult to Injury AKI Study reported in 2009 that only 50% of patients who died from a diagnosis of AKI received good care [17]. We feel probable that a suboptimal management of patients who developed AKI could be one of the reasons why CRRT effectiveness profile worsened after ICU discharge. 

Our study has several limits, in addition to the obvious ones related to its retrospective design. It was not possible to carry out post-hoc analysis because the patients’ data extracted from the “GiViTI” database are aggregated. Nevertheless, we demonstrated that the difference in mortality during and after the ICU stay is significant.

In conclusion, AKI is a clinical condition whose treatment requires a significant use of resources in the Italian ICUs and hospitals. Moreover, 30 days after the ICU discharge the mortality rate of the study population increased and significantly exceeded the SOFA Score prediction. Therefore, as the human and wealth capital spent for CRRT is quite conspicuous, we suggest developing a specific follow-up program for patients with an increased mortality risk due to AKI. The identification of all these patients, who need specific care to help them “see the light”, may, at the same time, prevent enlarging the hospital “black hole” of negative outcome and wasteful use of resources.

Tables and Figures

Table 1

Table 2

Figure 1

Figure 2

The authors declare that they have no conflict of interest

No funding sources

The study was approved by the University of Perugia

1. Uchino, Shigehiko, et al. "Acute renal failure in critically ill patients: a multinational, multicenter study." Jama 294.7 (2005): 813-818. doi:10.1001/jama.294.7.813

2. Klarenbach, Scott,  et al. "Economic evaluation of continuous renal replacement therapy in acute renal failure." International journal of technology assessment in health care 25.3 (2009): 331-338.DOI: https://doi.org/10.1017/S0266462309990134

3. Vincent, Jean-Louis, et al. "Use of the SOFA score to assess the incidence of organ dysfunction/failure in intensive care units: results of a multicenter, prospective study." Critical care medicine 26.11 (1998): 1793-1800. https://journals.lww.com/ccmjournal/abstract/1998/11000/use_of_the_sofa_score_to_assess_the_incidence_of.16.aspx

4. Goeree, R., & Diaby, V. (2013). Introduction to health economics and decision-making: Is economics relevant for the frontline clinician?. Best Practice & Research Clinical Gastroenterology, 27(6), 831-844. Goeree, Ron, and Vakaramoko Diaby. "Introduction to health economics and decision-making: Is economics relevant for the frontline clinician?." Best practice & research Clinical gastroenterology 27.6 (2013): 831-844.https://doi.org/10.1016/j.bpg.2013.08.016

5. Hoyt, D. B. (1997). CRRT in the area of cost containment: Is it justified?. American journal of kidney diseases, 30(5), S102-S104. Hoyt, David B. "CRRT in the area of cost containment: is it justified?." American journal of kidney diseases 30.5 (1997): S102-S104. https://doi.org/10.1016/S0272-6386(97)90550-3

6. Guidet, B., & Beale, R. (2015). «Should cost consideration be included in medical decisions? Yes,» Intensive Care Med, n. 41, 1839-1840. Guidet, B., and R. Beale. "Should cost considerations be included in medical decisions? Yes." Intensive Care Medicine 41 (2015): 1838-1840. DOIhttps://doi.org/10.1007/s00134-015-3988-6

7. "Kidney disease: improving global outcomes (KDIGO) acute kidney injury work group. KDIGO clinical practice guideline for acute kidney injury." Kidney international supplements 2.1 (2012): 1-138.https://doi.org/10.1038/kisup.2012.1

8. Vincent, Jean-Louis, et al. "Use of the SOFA score to assess the incidence of organ dysfunction/failure in intensive care units: results of a multicenter, prospective study." Critical care medicine 26.11 (1998): 1793-1800.https://journals.lww.com/ccmjournal/abstract/1998/11000/use_of_the_sofa_score_to_assess_the_incidence_of.16.aspx

9. GraphPad Software Inc. (2016). «Prism 7.02 for WIndows,» La Jolla, California.

10. Minne, Lilian, Ameen Abu-Hanna, and Evert de Jonge. "Evaluation of SOFA-based models for predicting mortality in the ICU: A systematic review." Critical care 12.6 (2008): 1-13.https://doi.org/10.1186/cc7160

11. Matsumura, Yosuke, et al. "Serum procalcitonin level and SOFA score at discharge from the intensive care unit predict post-intensive care unit mortality: a prospective study." PLoS One 9.12 (2014): e114007.

12. Mazzola, Paolo, et al. "The sequential organ failure assessment score predicts 30-day mortality in a geriatric acute care setting." Journals of Gerontology Series A: Biomedical Sciences and Medical Sciences 68.10 (2013): 1291-1295.DOI: https://doi.org/10.1093/gerona/glt020

13. Chang, Chih-Hsiang, et al. "Acute kidney injury enhances outcome prediction ability of sequential organ failure assessment score in critically ill patients." PloS one 9.10 (2014): e109649. DOI: https://doi.org/10.1371/journal.pone.0109649

14. Rizal, A. M., et al. "Cost-effectiveness analysis of continuous venovenous hemofiltration therapy in the treatment of sepsis in hospital Universiti Kebangsaan Malaysia: a comparison between high volume and standard rose hemofiltration." Jurnal Kesihatan Masyarakat (Malaysia) 12.1 (2006).https://journalarticle.ukm.my/4460/1/Vol12(1)-Rizal.pdf

15. Fagugli, Riccardo Maria, et al. "Six-year single-center survey on AKI requiring renal replacement therapy: epidemiology and health care organization aspects." Journal of nephrology 28 (2015): 339-349. DOIhttps://doi.org/10.1007/s40620-014-0114-8

16. Abraham, Kottarathil A., et al. "Inequalities in outcomes of acute kidney injury in England." QJM: An International Journal of Medicine 105.8 (2012): 729-740. https://doi.org/10.1093/qjmed/hcs037

17. Stewart, J. A. D. "Adding insult to injury: care of patients with acute kidney injury." British Journal of Hospital Medicine (2005) 70.7 (2009): 372-373. https://doi.org/10.12968/hmed.2009.70.7.43116