Prevalence & risk factors of multi-morbidity in critically ill patients with sepsis-associated acute kidney injury (SA-AKI)

Materials & Methods

This was a cross-sectional study conducted at the department of Critical Care Medicine, Yashoda Hospital, Hyderabad, Telangana, India, in adult patients aged 18 yr and above, admitted with sepsis to the ICU from January 2023 to January 2024. The study proceeded after obtaining ethics clearance from the institutional ethics committee. We included patients who stayed in the ICU for at least 72 h and developed AKI within seven days of admission (SA-AKI). The least serum creatinine value during the hospital admission was taken as baseline creatinine, and the stages of the AKI were defined based on the Acute Kidney Injury Network (AKIN) criteria⁵. Purposive sampling was used to select ICU patients diagnosed with SA-AKI, ensuring the sample was relevant to the study’s focus on multi-morbidity and its impact on outcomes. This method allowed for targeted data collection from individuals who met specific clinical criteria, enhancing the study’s relevance. Patient data was collected after obtaining their signed informed consent. Re-admissions and the patients prognosed to have chronic kidney disease (CKD) were excluded from the study to ensure that the results specifically reflect the acute impact of sepsis on kidney function, rather than the long-term effects of pre-existing kidney conditions. CKD patients often have a baseline decline in kidney function, which could confound the findings related to SA-AKI, as their renal impairment may not be solely attributed to the acute infection. Additionally, re-admissions could introduce bias, as these patients may have different clinical histories or complications that could skew the study’s focus on the initial presentation and progression of SA-AKI.

Results

In our cohort of 185 patients diagnosed with SA-AKI, the median age was 60 yr (IQR: 48, 70), with male predominance of 55.7 per cent (n=103). The BMI categorisation revealed that 57.8 per cent (n=107) had abnormal weight. The majority of patients were in stage II AKI (n=76, 41.1%; Table I). The primary causes of SA-AKI were classified into several categories: acute tubular necrosis accounted for 37.3 per cent (n=69), pre-renal factors represented 32.4 per cent (n=60), multifactorial causes comprised 24.3 per cent (n=45), nephrotoxicity was observed in 3.2 per cent (n=6), and hepato-renal causes were identified in 2.7 per cent (n=5). The most common sources of infection were urinary tract (32.4%, n=60), followed by gastrointestinal system (18.4%, n=34), lungs (15.1%, n=28), and blood stream (14.1%, n=26).

Electrolyte imbalances were frequently observed among participants, with hyponatremia, hypokalaemia, hyperchloremia, and hypoalbuminemia affecting 34.1 per cent (n=63), 25.9 per cent (n=48), 36.7 per cent (n=66), and 77.8 per cent (n=144) of patients, respectively. Elevated serum creatinine levels were noted in 91.9 per cent (n=170), with no patients showing low levels (Figure).

Close

Figure.

Distribution of serum electrolyte levels among the study population. Light grey represents hypo-electrolytemia, medium grey represents normal electrolytemia and dark grey represents hyper-electrolytemia)

Export to PPT

Diabetes mellitus and hypertension were notably prevalent, with 91.7 per cent (n=66) and 98.6 per cent (n=71) of multi-morbid patients affected, while their rates among overall SA-AKI patients were 46.5 per cent (n=86) and 52.4 per cent (n=97), respectively. In contrast, coronary artery disease was present in 22.2 per cent (n=16) of multi-morbidity patients and 9.7 per cent of those with SA-AKI, while chronic liver disease and cerebrovascular accidents were less common, affecting only 4.2 per cent and 8.3 per cent of multi-morbid patients, respectively.

The prevalence of multimorbidity was found to be 38.9 per cent (n=72). The association of various risk factors and presence of MM, highlighting the statistically significant relationships were displayed in table II. The regression analysis revealed that age, chloride imbalance, and albumin imbalance were significantly associated with multimorbidity. Individuals aged 40-59 yr and 60 yr and above had higher odds of multi-morbidity compared to those aged 18-39 yr, with the association becoming stronger in older age groups [Adjusted odds ratio (AOR) for 40-59 yr: 7.25, P=0.018; AOR for 60 yr and above: 22.11, P<0.001]. Those with chloride imbalance and albumin imbalance were less likely to be associated with multimorbidity, with significant AOR values of 0.42 and 0.10, respectively. Conversely, while stage 3 AKI initially showed lower odds of multi-morbidity (UOR: 0.35, P=0.012), this association weakened after adjustment (AOR: 0.41, P=0.061), and stage 2 AKI showed no significant association with multi-morbidity. Other factors like gender, BMI, and potassium imbalance did not demonstrate consistent associations with MM (Table III).

Among participants with multi-morbidity, 62.5 per cent (n=45) required mechanical ventilation, compared to only 28.3 per cent (n=32) of those without. The vast majority of both groups did not require renal replacement therapy, with 95.8 per cent (n=69) of multi-morbid patients and 95.6 per cent (n=108) of non-multi-morbid patients. All-cause mortality during hospital stays was significantly higher among patients with multi-morbidity (60.6%; n=43; Table IV).

Discussion

The prevalence of multi-morbidity among patients diagnosed with SA-AKI in our study was found to be 38.9 per cent, highlighting a significant burden of chronic health conditions in this population. Studying the risk factors and role of multi-morbidity in influencing outcomes among critically ill patients, especially with sepsis, is crucial in the management of SA-AKI. Failure of which may worsen the condition, leading to multiple organ dysfunction or failure and even death.

The demographic analysis of our participants revealed a broad age range, with significant findings related to the association between age and multimorbidity. The majority of patients were aged between 48 and 77 years, with a notable increase in multi-morbidity among those aged 63 to 77 years, with 45.8 per cent being multi-morbid. Like other studies investigating multi-morbidity among varied diagnoses, we observed multi-morbidity increasing with age, demonstrating a positive association^16-19. Older patients often had a greater burden of chronic diseases, which could contribute to the severity of acute conditions like SA-AKI.

Our findings also revealed a male predominance (55.7%) in the cohort. Although this difference was not statistically significant, it reflected a trend noted in other studies where male patients often presented with more severe comorbidities and worse outcomes^19-21. It is essential to consider gender differences in multi-morbidity when developing care strategies, as these may influence both the presentation and management of SA-AKI²¹.

Obese patients are prone to multiple chronic health conditions, including an increase in the prevalence of diabetes²². Our study focused on the co-occurrence of morbidities, and thereby, the effect these diseases could have on clinical outcomes when combined rather than individually. Diabetes mellitus and hypertension were prevalent among patients with multi-morbidity, affecting 91.7 per cent and 98.6 per cent of this group, respectively, which was similar to the studies in which hypertension and DM were documented as the most common comorbidities among sepsis patients^17,23,24. While studies conducted in Palestine, Middle East Asia, and Saudi Arabia reported CVD as common^19,21,25.

Our study found that 76.4 per cent (n=55) of patients had at least two comorbidities, followed by 22.2 per cent and 1.4 per cent with three and four comorbidities, respectively. Such proportions were higher than those documented from Nepal (12.9%) and Jordan (62.2%)^26-28.

The association between diabetes, hypertension, and kidney function deterioration is well-documented, underscoring the need for monitoring and managing these chronic conditions in at-risk populations. The high rates of these comorbidities in our cohort suggested that patients with SA-AKI often had multiple health challenges that might have complicated their clinical management. 

Electrolyte imbalances were prevalent among our participants, with hyponatremia and hypoalbuminemia being particularly common in the multi-morbid group. This finding was slightly different from a study conducted in Kathmandu with hypokalaemia in 8.99 per cent among 59.7 per cent sepsis-associated AKI patients²⁶; further emphasising the need for nutritional and metabolic monitoring in this population.

The sources of infection among our study participants were quite significant, with urinary tract infections being the most prevalent, accounting for 32.4 per cent of cases, followed by gastrointestinal (18.4%) and lung infections (15.1%). This finding was similar to other studies^26,29,30. The significant association between specific infection types and multimorbidity, particularly lung and skin infections, suggests that patients with multiple comorbidities may have had a higher susceptibility to severe infections, complicating their clinical management.

Among the multimorbid SA-AKI patients, 62.5 per cent required mechanical ventilation, compared to only 28.3 per cent of those without multimorbidity. This aligns with the findings of a study conducted in other parts of India²⁷ that showed significantly higher ventilator requirement rates compared to studies from other geographical settings such as Egypt, Europe and Nepal^19,26,31.

Interestingly, while the need for renal replacement therapy (RRT) did not show a significant difference between the groups, the majority of both multi-morbid and non-multi-morbid patients did not require this treatment. This finding suggests that, while multi-morbidity increases the complexity of care, it may not necessarily translate to an increased need for RRT in all cases. Future studies could explore this relationship further to clarify the implications of multi-morbidity on renal management strategies. RRT had a significant association with sepsis mortality, aligning with the literature³².

The all-cause mortality rate during hospital stays was significantly higher among patients with MM, with 60.6 per cent; similar to the findings of 32.2 to 75 per cent ICU mortality in varied geographical settings^23,30,33. Literature also suggests that an increase in the number of comorbidities increases the odds of death²³.

This study had several strengths, including its focus on the critical issue of SA-AKI and the underexplored role of MM in critically ill patients, particularly in LMICs like India. It identifies key risk factors associated with poor outcomes, highlighting the clinical relevance of addressing both acute and chronic health conditions in this high-risk population. Additionally, the study provided real-world insights into the prevalence of multi-morbidity and its association with increased mortality, mechanical ventilation needs, and complications in SA-AKI patients, contributing valuable information for improving patient care and clinical management strategies. By exploring these relationships in the context of sepsis and kidney injury, the study laid the groundwork for future research.

However, our investigation had a few limitations. The cross-sectional study design limited the ability to establish causal relationships between MM and SA-AKI outcomes. A longitudinal study may help develop a deeper understanding of such associations. Furthermore, there may be unmeasured confounding factors, like genetic predispositions or undetected comorbidities, which could have influenced the outcomes. Participants were selected using purposive sampling due to operational and logistical constraints. While purposive sampling allowed for targeted data collection, we acknowledge that a consecutive sampling approach would have reduced potential selection bias and improved generalisability. Lastly, the accuracy of the results depended on the completeness and reliability of the medical records, which could introduce potential reporting errors or omissions. While missing or incomplete data were minimal (less than 5%), particularly in laboratory results and documentation of comorbidities, we acknowledge that such omissions could have a small impact on the results.