Patient selection

Between 2009 and 2018, we reviewed 628 patients with an acute febrile illness and a rash who were diagnosed with scrub typhus confirmed by a positive IgM ELISA (enzyme-linked immunosorbent assay) (InBios International Inc., Seattle, WA) for scrub typhus. Patients who were transferred to another hospital for a higher level of care, or to be treated nearer to their own homes during the course of treatment or had concomitant infections like leptospirosis, malaria, or dengue fever were excluded from the study. In addition, patients who were not admitted to our hospital were excluded from this study. We also excluded patients who were not followed-up during the complete recovery in their renal function or for at least three months after discharge. However, we included patients who were transferred from a local hospital after their baseline renal function had been checked. Therefore, a total of 611 patients were enrolled in this study and were divided into two groups [normoalbuminemia (NA) vs. HA] based on the serum albumin level of 3.0 g/dL. This study was approved by the Institutional Review Board of the Presbyterian Medical Center, Jeonju, South Korea (IRB No. 2020-04-017).

Clinical and laboratory information

The patients’ baseline demographic, clinical, and laboratory data were reviewed at the time of hospitalization and during the follow-up period. Detailed clinical histories of all patients were determined and they underwent thorough physical and biochemical examinations. A standard set of investigations including the following were carried out: complete blood counts, liver function and renal function tests, chest radiograph, three peripheral blood smears for malaria, urinalysis (including urea and electrolytes), two blood cultures, and a standard set of febrile serological investigations including an ELISA for the detection of scrub typhus (InBios International Inc., Seattle, WA). To differentiate between endemic zoonoses including leptospirosis and hemorrhagic fever with renal syndrome (HFRS), another serological work-up was performed. Leptospirosis was diagnosed by a ≥ 4-fold rise in indirect immunofluorescent assay (IFA) titer in paired serum samples, or an antibody titer ≥ 1:800 in one serum sample by the microscopic agglutination test. HFRS was diagnosed by a single titer ≥1:80 or ≥ 4-fold rise in IFA titer in paired serum samples.

The complications were defined based on the presence of dysfunction of the following organ systems: 1) Cardiovascular system—presence of any of the following: systolic blood pressure <90 mmHg, myocarditis (defined using the diagnostic criteria for clinically suspected myocarditis presented by the European Society of Cardiology [16]), or new-onset cardiac arrhythmia including atrial fibrillation or supraventricular tachycardia; 2) Respiratory system—presence of any of the following: acute respiratory failure (defined not only by the incidence of hypoxemia, hypercapnia, or the failure to maintain both within a normal range, but also by the presence of any condition that resulted in the need for mechanical ventilation), pneumonia, pleural effusion, pulmonary edema, focal atelectasis, patchy consolidation on chest radiograph or acute distress syndrome; 3) Central nervous system—presence of any of the following: Glasgow Coma Scale ≤12 in the absence of any underlying causes, seizure without underlying causes, meningitis (defined by the presence of headache or nuchal rigidity with either altered sensorium or focal neurological deficits on history or examination, and with cerebrospinal fluid cell counts ≥5 leukocytes/mm³); 4) Urinary system—presence of AKI, defined based on the RIFLE (Risk, Injury, Failure, Loss of kidney function, and End-stage kidney disease) criteria; 5) Gastrointestinal and hepatobiliary system—presence of any of the following: presence of hepatitis (defined as an elevation of serum transaminases more than five times the normal upper limit), or hyperbilirubinemia as serum total bilirubin >2 mg/dL, or gastrointestinal bleeding [9, 17–19].

Patients with AKI were categorized into the Risk (R), Injury (I) and, Failure (F) [20]. The estimated glomerular filtration rate (eGFR) was estimated using the abbreviated Modification of Diet in Renal Disease (MDRD) equation [21]. When the baseline renal function was not obtainable, it was calculated using the standard four-variable MDRD formula, assuming an eGFR of 75 mL/min/1.73 m². The RIFLE class was determined according to the worst of either the serum creatinine, or eGFR, and urine output criteria. Renal replacement therapy including intermittent hemodialysis was initiated using standard indications. All data are presented as mean ± standard deviation unless otherwise specified. The baseline characteristics of patients in the two groups were compared using t-tests, the chi-square test, or Fisher’s exact test, as appropriate. The clinically relevant parameters or the variables that were significantly associated with the presence of AKI in the univariate analysis were included in the multivariate analysis. The AKI group was further grouped into three subgroups, then serum albumin level was compared among the three subgroups by the one-way analysis of variance followed by post Hoc analysis. A p-value of <0.05 was considered statistically significant. Statistical analysis was carried out using SPSS version 22.0 (IBM Corp., Armonk, NY).

Comparison of clinical characteristics between the NA and HA groups

Compared to the patients in NA group, the patients in the HA group were older (73 ± 9 vs. 62 ± 14 years, P<0.001) and had a higher total leukocyte count (9.6 × 10³/ mL vs. 6.2 × 10³/mL, P<0.001) (Table 1).

Table 1

Comparison of baseline characteristics between NA and HA groups.

	NA	HA	P-value
	(n = 533)	(n = 78)
Age	62 ± 14	73 ± 9	< 0.001
Male, n(%)	241 (45)	30 (39)	NS
Duration of hospital stay, days	6.2 ± 3.1	9.6 ± 6.2	< 0.001
DM, n(%)	90 (17)	19 (24)	NS
Hypertension, n(%)	188 (35)	40 (51)	0.005
CKD, n(%)	19 (3)	3 (4)	NS
ICU care, n(%)	11 (2.1)	16 (20.5)	< 0.001
Hemoglobin (mg/dl)	12.8 ± 1.6	11.5 ± 1.6	< 0.001
Complications
Gastrointestinal tract, n (%)	68 (13)	8 (10)	NS
Respiratory system	73 (14)	39 (50)	< 0.001
Cardiovascular system	58 (11)	22 (28)	< 0.001
Central nervous system	4 (0.8)	5 (6.4)	0.002
Acute kidney injury	96 (18)	45 (58)	< 0.001
Leukocyte (×103/ mL)	6.7 ± 4.1	10.2 ± 3.8	< 0.001
Platelet count (×103/ mL)	141± 57	122 ± 48	0.006
Total bilirubin level	0.7 ± 0.4	0.7 ± 0.7	NS
Serum albumin (mg/dl)	3.7 ± 0.4	2.7 ± 0.2	< 0.001
Serum ALT (IU/L)	91 ± 129	76 ± 56	NS
Creatinine (mg/dl)	1.0 ± 0.4	1.3 ± 0.8	< 0.001
eGFR ml/min/1.73m2	71 ± 24	50 ± 25	< 0.001

The duration of the hospital stay was longer for patients in HA group than NA group (9.6 ± 6.2 vs 6.2 ± 3.1 days, p<0.001). Furthermore, the patients in HA group required ICU care more frequently than those in NA group (20.5% vs 2.1%, P<0.001). The HA group also had a higher incidence of complications in the respiratory (50% vs 14%, P<0.001), cardiovascular (28% vs 11%, P<0.001), and neurologic systems (6.4% vs 0.8%, P = 0.002). In addition, AKI (58% vs 18%, P<0.01) developed more frequently in HA group than in NA group. The plasma ALT concentrations did not differ between the two groups, whereas the eGFR on admission (50 ± 25 ml/min/1.73m² vs. 71 ± 24 ml/min/1.73m², P<0.001) was lower in the HA group than in NA group.

Comparison of clinical characteristics between the non-AKI group and AKI group

Compared to patients in the non-AKI group, the patients in the AKI group were older (71 ± 11 vs. 62 ± 14 years, P<0.001) and had a higher incidence of comorbidities, such as hypertension, diabetes, or chronic kidney disease (Table 2).

Table 2

Comparison of baseline characteristics between AKI and non-AKI group.

	Non-AKI	AKI	P-value
	(n = 470)	(n = 141)
Age	62 ± 14	71 ± 11	< 0.001
Male, n(%)	200 (43)	71 (50)	NS
Duration of hospital stay, days	6.1 ± 2.7	8.4 ± 5.7	< 0.001
DM, n(%)	66 (14)	43 (31)	< 0.001
Hypertension, n(%)	138 (29)	90 (64)	< 0.001
CKD, n(%)	5 (1)	17 (12)	< 0.001
Hemoglobin (mg/dl)	12.9 ± 1.6	12.1 ± 1.9	< 0.001
Leukocyte (×103/ mL)	6.5 ± 4.0	9.3 ± 4.3	< 0.001
Platelet count (×103/ mL)	143 ± 57	121 ± 53	< 0.001
Total bilirubin level (mg/dl)	0.7 ± 0.3	0.8 ± 0.6	NS
Serum albumin (mg/dl)	3.7 ± 0.5	3.2 ± 0.5	< 0.001
Serum ALT (IU/L)	90 ± 119	83 ± 133	NS
Creatinine (mg/dl)	0.8 ± 0.2	1.5 ± 0.7	< 0.001
eGFRml/min/1.73m2	78 ± 20	39 ± 17	< 0.001

In addition, patients in the AKI group had poorer renal function (39 ± 17 vs. 78 ± 20 mL/min/1.73m², P<0.01) on admission and had a higher total leukocyte count (10.01 × 10³/ mL vs. 6.78 × 10³/mL, P<0.01). The plasma ALT concentrations did not differ between the two groups. However, the serum albumin level (3.2 ± 0.5 mg/dL vs. 3.7 ± 0.5 mg/dL, P<0.001) was lower in the AKI group than that of the non-AKI group.

The relationship between serum albumin and acute kidney injury

Among patients with AKI, 15 (11%) were oliguric (Table 3).

Table 3

Clinical characteristics of 141 patients with AKI.

Non-oliguric	15 (11)
FENa < 1%, n (%)a	53 (64)
Recovery of renal function within 72 h, n (%)	138 (98)
Renal function
eGFR adm, ml/min/1.73m2	39 ± 17
eGFR low, ml/min/1.73m2	36 ± 16
eGFR rec, ml/min/1.73m2	74 ± 21
Urinalysis results
Proteinuria	15 (32)
Pyuria	5 (10)
Hematuria	1 (2)
Normal	26 (56)

FENa: fractional excretion of sodium

^a FENa was available in 83 patients.

eGFR adm, ml/min/1.73m^2: eGFR at the time of admission

eGFR low, ml/min/1.73m^2: eGFR at lowest value in hospital

eGFR rec, ml/min/1.73m^2: eGFR at the time of recovery

One hundred and thirty-five patients had AKI prior to admission and 6 patients experienced AKI during their hospitalization. Of the 141 patients with AKI, 133 experienced a return to their baseline renal function within 72 hours after admission. Of the 83 patients with available fractional excretion of sodium (FENa) data, 53 (64%) had a FENa < 1%. Applying the RIFLE criteria, 91 (64.5%), 43 (30.5%), and 7 (5.0%) patients were classified into the R, I, and F categories, respectively. The mean serum albumin concentrations differed among categories (Risk: 3.4 ± 0.5 mg/dL vs. Injury: 3.0 ± 0.5 mg/dL vs. Failure: 2.6 ± 0.3 mg/dL, P<0.005) (Fig 1).

Fig 1

The mean serum albumin concentrations among categories in patients with AKI.

The serum albumin level correlated with AKI severity. ^ap<0.05 vs. Risk, ^bp<0.05 vs. Risk, ^cp<0.05 vs. Injury.

On univariate analysis, age, presence of comorbidities, such as hypertension, diabetes, or chronic kidney disease, hemoglobin, thrombocytopenia, total leukocyte count, and hypoalbuminemia were significant predictors of AKI. After adjusting for these factors in multivariate logistic regression analysis, the presence of comorbidities, leukocytosis, and hypoalbuminemia were significant predictors of AKI (Table 4).

Table 4

Predictors of AKI (univariative and multivariative analysis).

	Univariate		Multivariate
	HR (95% CI)	p-value	HR (95% CI)	p-value
Age	1.054 (1.036–1.072)	<0.001	1.019 (0.999–1.040)	0.064
DM	2.686 (1.725–4.183)	<0.001	2.037 (1.179–3.519)	0.011
Hypertension	4.246 (2.855–6.313)	<0.001	2.337 (1.449–3.769)	0.001
CKD	12.750 (4.613–35.239)	<0.001	8.265 (2.868–28.528)	<0.001
Hemoglobin	1.324 (1.178–1.488)	0.703	1.003 (0.866–1.162)	0.963
Leukocytosis	1.179 (1.120–1.242)	<0.001	1.130 (1.069–1.195)	<0.001
Platelet	1.006 (1.002–1.009)	<0.001	1.005 (1.000–1.009)	0.051
Hypoalbuminemia	5.239 (3.469–7.912)	<0.001	2.925 (1.722–4.967)	<0.001