Prevalence, incidence, and treatment of anaemia in patients with non-dialysis-dependent chronic kidney disease: findings from a retrospective real-world study in Italy

To date, this is the largest real-world study on anaemia of NDD-CKD in Italy. Starting with records for approximately 1.5 million inhabitants from a pool of five areas geographically distributed across Italy, 101,143 patients with NDD-CKD were identified. Anaemia of NDD-CKD was characterised in terms of epidemiology, demographics, clinical characteristics, and management.

The overall standardised prevalence of NDD-CKD in 2016 was 5.6%. Previous Italian studies describing prevalence rates for CKD have not reported data on anaemia, or focused on NDD-CKD, and used different patient inclusion criteria; therefore, those findings are not directly comparable with the results from this research [2, 17]. In this study, the standardised prevalence rate in 2016 for anaemia among patients with NDD-CKD was 33.8%, which differs from rates (41–62%) reported by other Italian studies enrolling patients regularly followed in renal clinics [11, 12]. Different prevalence rates reported by studies are expected given the variations in study populations and methodology. Patients with NDD-CKD were predominantly female in the present study. This is consistent with previous studies, which report that the prevalence of CKD was higher in female patients, despite increased severity and risk of progression of CKD in male patients [18]. In the present study, the prevalence of anaemia among patients with NDD-CKD was numerically higher in males than females, which conflicts with findings from previous studies that females were more likely to develop anaemia of CKD than males [3, 19]. This may be a consequence of the higher Hb cut-off value in male patients compared to females (< 13 g/dL versus < 12 g/dL) for the definition of anaemia in the present study. As a result, the Hb cut-off value might be attained at an earlier stage of disease in male than female patients [20]. Furthermore, comorbidities with the potential to induce anaemia (e.g., CV disease and diabetes) were reported in higher proportions of male patients than female patients in this study. Standardised prevalence and incidence of anaemia among patients with NDD-CKD increased through CKD stages, suggesting a greater need for anaemia management in patients with more advanced CKD. Trends in the standardized incidence of anaemia among patients with NDD-CKD from 2014 to 2016, overall and by CKD stage, were variable; particularly for stage 3a where anaemia was less frequent.

Anaemia is known to become more common as CKD progresses, and may be associated with increased risks of cardiovascular disease, hospitalisation, cognitive impairment, and death [6]. In the current study, hypertension was the most frequent comorbidity, with rates of 83.8% to 96.4% across the three cohorts, and without a clear relationship with CKD stage. These observations are in line with other studies in patients with NDD-CKD, in which about 80–87% of patients had hypertension [21, 22]. Similarly, no trend across CKD stages was observed for the prevalence of diabetes mellitus, a clinical condition more frequently associated with new-onset anaemia [23]. The present study also showed that a diagnosis of anaemia in patients with CKD was associated with a higher risk of ESRD, and of death, independently of risk factors related to adverse outcomes in this patient population [11, 22, 24, 25].

Among anaemic patients, we applied two different cut-offs of Hb concentration (≥ 2 records of either Hb < 10 or < 11 g/dL over 6 months) in estimating ESA eligibility to reflect international guidelines and clinical practice in Italy, respectively. This ensured that the data would be relevant worldwide and not just in Italy. Requiring ≥ 2 records over 6 months rather than a single record reduced the chance that a reduction in Hb may have had a temporary cause such as inflammation, infection, bleeding, or surgery [26]. Based on the two eligibility criteria, 44.2% (Hb < 10 g/dL) and 63.4% (Hb < 11 g/dL) of anaemic patients were eligible for ESA treatment. Regardless of the Hb cut-off, the proportion of eligible patients actually treated with ESAs was low (≤ 15.5%), although the proportion treated increased across the CKD stages (from < 10% in stage 3a to approximately 40% in stage 5). These results suggest that patients who required treatment for anaemia were inadequately treated when ESAs could have been prescribed. Therapeutic inertia has been reported previously in the context of nephrology care, with 34% of NDD-CKD patients not receiving ESAs despite having anaemia over a 6-month observation period [12]. This phenomenon has been confirmed recently by the CKDopps analysis, which showed that a large proportion of patients with anaemia of NDD-CKD did not receive anaemia medication within 1 year [10]. However, it would be an oversimplification to conclude from the present study that anaemia was undertreated. Indeed, international guidelines recommend that the decision to treat with ESAs is not based on Hb level alone, but should also consider symptoms related to anaemia, prior cardiovascular history (e.g., stroke), active or past history of malignancy, rate of decrease of Hb concentration, prior response to iron therapy, the risk for transfusion, and risks related to ESA therapy [7, 26]. Therefore, the large proportion of untreated patients may reflect the need for healthcare professionals to understand the underlying causes of anaemia before prescribing ESAs, while also reducing the burden of cost of inappropriate prescriptions on the National Health Service. In addition, this large untreated patient population may rely on a sub-optimal healthcare service, which may not refer all patients to the appropriate specialist, such as a nephrologist, to receive treatment. Of note, the present analysis was based on ESAs prescribed to patients in the outpatient setting, and ESA use during hospitalisation was not captured. However, ESAs prescribed during hospitalisation continue to be given after discharge, and therefore the absence of hospital records of ESA use is unlikely to have affected the study findings.

The proportion of anaemic patients who received IV iron infusions in the present study (9.0%) was generally consistent with data reported in other studies. A study using data from the CKDopps found that IV iron was received by 12%, 9%, 33%, and 8% of patients with anaemia of CKD stages 3a–5 in Brazil, France, Germany, and the USA, respectively [9]. Another study conducted in Italy found that IV iron was prescribed to approximately 3% of NDD-CKD patients receiving iron supplements [12]. A challenge remains in the estimation of the proportion of patients receiving oral iron therapy. Oral iron supplements are likely to be the predominant form of iron therapy in patients with NDD-CKD, but their over-the-counter availability makes their use more difficult to track. As oral iron data were not readily available in the databases used for the current study, these data were not analysed for the follow-up period.

An original finding of this study was the rate of blood transfusions in the NDD-CKD population. Studies from the USA have reported a rising prevalence of blood transfusions in the non-dialysis population, which has been linked with a concomitant and marked reduction in the use of ESAs [27]. This greater use of blood transfusions than ESAs or IV iron is believed to have been triggered by policy changes, including drug reimbursement, and additional FDA ‘black box’ warnings related to the potential adverse effects of ESAs [27]. Indeed, in one US study, the proportion of NDD-CKD patients aged 66–85 years with anaemia who received blood transfusions (22.2%) was found to be almost double the proportion observed in anaemic patients (11.3%) in the present study [27]. This finding could be explained by differences between the US and Italian healthcare systems. Of note, in this study the use of blood transfusions decreased with advancing CKD stage in parallel with increasing ESA prescriptions. This trend was independent of the severity of anaemia (either Hb < 10 g/dL or < 11 g/dL).

The ‘real-world’ setting and large sample size are major strengths of this study. Using population-based data, a ‘snapshot’ has been provided of the management of anaemia in more than 100,000 patients with NDD-CKD in Italy, a population for whom these epidemiologic data were lacking. This study, however, has limitations. We cannot exclude that other haemoglobinopathies (in particular thalassaemia) have been classified as renal anaemia in our study, although the advanced age of our study population strongly limits the potential impact of thalassaemia on our estimates. Also, the inclusion of patients with anaemia from non-CKD-related causes in this study may have resulted in a slight overestimation of the number of patients eligible for ESAs, and the occurrence of clinical inertia. Frequently, the use of large administrative databases to investigate a specific clinical question can be associated with limited suitable data, thus reducing cohort size [28]. The databases captured only direct medical healthcare resource use reimbursed by the Italian National Health Service. In patients without ICD-9-CM codes for CKD, only one creatinine value was used to define the presence of CKD, whereas international guidelines require at least two pathological creatinine values or other markers of kidney damage more than 3 months apart [29]. This discrepancy may have resulted in a partial overestimation of CKD cases. A proxy (e.g., specific drugs, hospitalisations) was used for comorbidities, such as diabetes, which may have introduced error in estimating their prevalence.

The length of look-back period can also impact estimates of prevalence and incidence: a shorter look-back period can potentially lead to the overestimation of an incidence rate due to the misclassification of prevalent and recurrent cases as incident cases [30]. The minimum look-back period for inclusion in this study was 1 year but could have been up to 8 years depending on the individual patient’s inclusion date; however, the length of look-back period was not analysed for this study. Other limitations include the evaluation of drug treatments based on prescriptions only, and therefore ESA prescriptions could have been related to conditions other than anaemia of CKD. Patients with a cancer diagnosis were excluded from the study in order to reduce this potential bias, as ESAs could be prescribed to these patients and therefore could have influenced the results for CKD and anaemia. Eligibility for ESA treatment was based on Hb levels alone. As other decision criteria for ESA treatment recommended by the guidelines, such as symptoms of anaemia and prior cardiovascular history, were not recorded in the LHU databases, it is not possible to comment on the extent to which patients with anaemia may have been undertreated [7, 26]. The iron supplementation data were limited to IV iron. Lastly, it is unknown to what extent the data for Italy from the current study may be generalised to other countries given the differences in healthcare systems.

In conclusion, this study demonstrated that anaemia is a significant issue in patients with NDD-CKD, and the incidence increases with CKD stage. The proportion of patients eligible for ESA treatment who actually received ESAs was relatively low, indicating a potential treatment gap, and suggesting that anaemia may not be adequately controlled. There appears to be an unmet need to remedy the apparent clinical inertia and improve the diagnosis and treatment of anaemia of NDD-CKD in Italy. However, this study also highlights the careful choices made by healthcare professionals when prescribing ESAs, accounting for both the underlying conditions of the patients, and the cost of inappropriate ESA prescription to the Italian National Health Service.

留言 (0)

沒有登入
gif