Comparison of Clinical Profile and in-Hospital Outcomes of Communityacquired Versus Hospital-Acquired Acute Kidney Injury in Hospitalized Patients in a Tertiary Care Hospital: A Prospective Study

ABSTRACT

BACKGROUND: Acute kidney injury (AKI) is a clinical syndrome denoted by an abrupt decline in glomerular filtration rate (GFR) sufficient to decrease the elimination of nitrogenous waste products (urea and creatinine) and other uremic toxins. Based on the type of setting AKI can be Community Acquired (CA-AKI) or Hospital Acquired (HA-AKI). These two types have different epidemiological etiological and outcome profiles and these characteristics have remained inconclusive. As far as the etiological spectrum is concerned;previous studies have demonstrated a varied spectrum in both these groups. Very few studies comparing the outcome of CA-AKI and HA-AKI were found in the literature search. There is a paucity of relevant comparative Indian studies on these two types of AKI. Hence this prospective observational study was undertaken to compare the demographic and clinical spectrum and short-term in-hospital outcomes of patients belonging to both these groups who were admitted to the largest tertiary care government teaching hospital in the state of Uttarakhand. AIM OF THE STUDY To compare the demographic and clinical spectrum and short-term in-hospital outcomes of community-acquired versus hospital-acquired Acute Kidney Injury in hospitalized patients METHODS: It is a prospective cohort study conducted from October 2020 to December 2021. The study was conducted in the In-Patient Department (IPD) areas of the Department of Nephrology and all those departments whose consultations for patients with suspected AKI were sent to the Department of Nephrology at AIIMS Rishikesh. Patients fulfilling the following inclusion criteria were enrolled in this study- Age -18 years and the patients diagnosed as having AKI as per KDIGO 2012 definition. Those aged <18 years of age and those with CKD or Acute on CKD were excluded from the study. CKD was defined as per the KDIGO 2012 definition. Each enrolled patient was classified as having Community-acquired AKI (CA-AKI) or Hospital-acquired AKI (HA-AKI). Those admitted to the hospital with AKI were denoted as having CA-AKI. In contrast, patients were identified as having HA-AKI when AKI was not apparent upon hospital admission but was diagnosed beyond 24 hours of hospitalization. The sample size of 65 in community-acquired AKI and 32 in the hospital-acquired AKI group was calculated. Study subjects underwent detailed history clinical examination and relevant investigations required in the management of AKI episodes. The stage of AKI at presentation was assessed as per KDIGO Clinical Practice Guidelines for Acute Kidney Injury 2012. Ethical clearance was obtained. RESULT(S) A total of 65 patients with CA-AKI and 32 patients with HA-AKI were enrolled. The mean age of patients in the CA-AKI group was 46.7 years and in the HA-AKI group was 45.5 years. The CA-AKI group had significantly higher-baseline serum creatinine (P < 0.001), serum creatinine at admission (P < 0.001), proportion of male patients (P = 0.09), proportion of patients requiring renal replacement therapy (P = 0.02), proportion of patients getting admitted to medical IPDs (P < 0.001), proportion of patients whose baseline creatinine was unknown (P < 0.001), proportion of patients presenting in Stage 3 of AKI (P = 0.001), proportion of patients having oligoanuria (P = 0.09) and hyperkalemia (P = 0.06) at presentation. The HA-AKI group, on the other hand, was found to have a significantly higher- proportion of patients getting admitted to surgical IPDs (P < 0.001), proportion of patients who underwent a prior surgical procedure (P < 0.001), proportion of patients having coexisting lung disease (P = 0.09), liver disease (P = 0.03), heart disease (P = 0.06) and COVID-19 (P = 0.04). Sepsis was found to be the most common cause (70.7%) in the CA-AKI group and was also one of the common causes (28.12%) in the HA-AKI group. Despite more patients in the CA-AKI group being in AKI-Stage 3 at presentation, in-hospital mortality was observed to be lower in this group (35.4% versus 62.5%, P = 0.04). The median survival time of patients was und to be more than double in the CA-AKI than in the HA-AKI group (59 days versus 23 days). However, on comparing the overall survival using the log-rank test, both groups were found to be comparable (chi-square value 1.82, p-value 0.18). Univariate analysis for predictors of mortality showed that the type of AKI (CA vs HA) (P = 0.01), type of admission (ward vs ICU) (P = 0.001), surgical procedure prior to AKI onset (P = 0.018), presence of comorbidities such as DM (P = 0.038), lung disease (P = 0.000), and COVID-19 (P = 0.018) and requirement of vasopressor support (P = 0.009) were significant predictors of mortality of patients with AKI admitted to our center. Also, the length of hospital stay (P = 0.037), serum creatinine at admission (P = 0.002) and serum creatinine at discharge/death (P = 0.003) have been found to predict the mortality of these patients. However, Cox proportional hazard regression analysis for finding out independent predictors of mortality showed that only two factors, i.e., the presence of lung disease (HR 2.65, 95% CI 1.03-6.79, P = 0.042) and the requirement of vasopressor support at presentation (HR 5.28, 95% CI 1.75- 15.97, P = 0.003) predicted the survival of AKI patients. Thus, the present study showed that type of AKI was not an independent predictor of mortality in AKI patients admitted to our center. CONCLUSION(S) The majority of patients in both groups of AKI presented in Stage 3. Sepsis was found to be the most common cause in the CA-AKI group and was also one of the common causes in the HA-AKI group. On comparing the inhospital outcomes of AKI episodes, it was observed that both recovery (complete or partial) and dialysis dependency were more common in the patients with CA-AKI while mortality was found to be more in the HA-AKI group. However, on Cox proportional hazard regression analysis it was found that only two factors, i.e., the presence of lung disease and the requirement of vasopressor support at presentation predicted the survival of AKI patients admitted to our center. Thus, the present study showed that type of AKI was not an independent predictor of mortality in such patients. Further, more long-term and larger multi-center studies are required to study the course and outcome of patients with AKI and to outline the regional variances in its patterns in the Indian population.