Community acquired and hospital acquired AKI - two diseases divided by a common definition

Abstract

Purpose of review

Acute kidney injury (AKI) is common across the world. AKI that is acquired in the community (community acquired AKI, CA-AKI) has different risk factors, epidemiological profile, presentation and impact as compared to hospital acquired AKI (HA-AKI). Hence, similar approaches to tackle CA-AKI and HA-AKI might not work. This review highlights the important differences between the two entities that have a bearing on the overall approach to the conditions and how CA-AKI has been overshadowed by HA-AKI in research, diagnosis and treatment recommendations and clinical practice guidelines.

Recent findings

The overall burden of AKI is disproportionately more in low and low-middle income countries. The Global Snapshot study of International Society of Nephrology’s (ISN) AKI 0by25 program has shown that CA-AKI is the dominant form in these settings. Its profile and outcomes vary with geographical and socio-economic characteristics of the regions where it develops. The current clinical practice guidelines for AKI align more with HA-AKI than CA-AKI, and fail to capture the complete spectrum of CA-AKI as well as its impact. The ISN AKI 0by25 studies have uncovered the circumstantial compulsions in defining and assessing AKI in these settings and shown feasibility of community-based interventions.

Summary

Efforts are needed to better understand CA-AKI in low-resource settings and develop context specific guidance and interventions. A multidisciplinary, collaborative approach with representation from community would be required.

Introduction

Acute deterioration in kidney function, called acute kidney injury (AKI), is a major organ illness that is associated with development of complications in other organ systems, prolonged hospitalization, high healthcare utilization and costs and increased mortality. The burden of AKI varies between different regions or settings worldwide. The difference in the epidemiology and outcomes of patients presenting to the hospital with established AKI developed outside a hospital setting and those who develop AKI in-hospital has been recognized for decades [1], and forms the basis for classifying AKI into community acquired AKI (CA-AKI) or hospital acquired AKI (HA-AKI). AKI which is present at the time of hospital admission or is identified soon after (typically within 48 h) hospital admission is more likely to have been initiated by factors prevalent in the community and hence known as CA-AKI. CA-AKI usually develops in the context of another illness or insult in otherwise healthy individuals. On the other hand, AKI that develops in hospitalized patients, usually after 48 h of hospital stay is likely related to diagnostic or therapeutic interventions during hospitalization or progression of complications due to underlying illnesses. Given that older patients or those with complex acute or chronic illnesses are at higher risk of hospitalization, such patients constitute a majority in HA-AKI datasets.

CA-AKI is reported most frequently from the resource constrained, developing countries located primarily in tropical or sub-tropical zones. The term ‘tropical AKI’ has also been used to described these conditions. Countries in these climatic zones comprise approximately 92% of the world’s inhabited land mass and 78% of the world’s population, yet account for just about 45% of the world’s gross national product [2]. There is a consistent increase in the per capita national gross domestic product of countries as their distance from the equator increases [3]. Only two small tropical regions (Singapore and Hong Kong) figure in the list of 32 countries classified as high-income by the World Bank. Importantly, tropical countries produce only 5% of global research output and are vastly under-represented in organizations and working groups that make clinical practice recommendations [4].

The tropical ecology, determined by local environmental and climatic conditions, favors biological abundance and high degree of diversity in plants and animals, and consequently persistence and propagation of disease-causing micro-organisms, vectors and non-human hosts. Agriculture is the largest livelihood generator in these countries, and most people must work outdoors in hot and humid external environment for long hours without adequate protection which increase chances of contact with these organisms. As anthropogenic climate change and human encroachment lead to degradation of the biodiversity, the likelihood of encroachment of human habitation by wildlife and vice versa is consistently increasing, leading to more frequent infectious disease outbreaks, many of which can lead to AKI. Clean drinking water is often scarce, especially for the poorer sections of the society, in many of these countries. Other public health problems such as poor obstetric care, poisonings, and animal bites and stings etc. also contribute to the continued high burden of CA-AKI. Lower socio-economic conditions and weak healthcare systems translate into delayed diagnosis and referrals, poorer access to healthcare and over-burdened healthcare systems. Over-reliance on alternative, traditional or unproven systems of medicine are additional barriers to prompt and appropriate care. Favorable geographical and socio-economic conditions, more organized and efficient primary healthcare systems and easier access to high quality medical care largely preclude exposure to the aforementioned community-based risk factors for AKI for a majority of the population in developed countries. In such circumstances, the dominant determinants of the risk of AKI, even when it develops outside hospitals, are patient’s own characteristics, for example, age, comorbidities, etc.

Most of the global burden of AKI is concentrated in low- and middle-income countries (L/MIC). According to a Systematic Review, an estimated 13.3 million cases of AKI are recorded worldwide every year, with developing countries contributing 11.3 million [6]. About three quarters of cases develop as a result of infections prevalent in the community, toxins (animal bites, herbs, and medications) and pregnancy complications [2,5]. Of the 1.7 million deaths per year from AKI globally, an estimated 1.4 million are in L/MICs [7].

Advances in the understanding of pathogenesis of AKIs have not translated into proportionally better ability to make an early diagnosis in those at risk or to offer tailored treatment approaches. AKI that develops after hospitalization almost always portends a poor prognosis irrespective of other factors or underlying clinical status. Most of the data regarding epidemiology, treatment, and outcomes in AKI come from large hospital-based cohorts in developed countries [8]. The resources for such data collection are either limited or not available in resource-constrained developing regions of the world. The few published studies show little difference between HA-AKI in developed and developing countries [9].

Acute Kidney Injury: the Importance of Context

Figure 1 depicts a contextual model of AKI where factors that impact AKI are grouped according to whether the patient is likely to experience these inside the hospital or in the community. With increasing standardization of diagnosis and management protocols, and implementation of evidence-based practices, hospital-based factors that associate with AKI are largely recognised as being similar in all parts of the world. However, the same is not true about community-based factors. In fact, it would not be an exaggeration to say that CA-AKI is more socially, economically and culturally driven whereas HA-AKI is driven more by individual and medical factors.

Figure 1. Contextual model of Acute Kidney Injury Community or Hospital based.

ARDS, acute respiratory distress syndrome; MODS, multiorgan dysfunction syndrome.

The Global Snapshot study of International Society of Nephrology’s (ISN) AKI 0by25 initiative recorded observations in >4000 patients with AKI across 72 countries and showed that CA-AKI was commoner than HA-AKI in low- and lower middle-income countries [10]. Also, it was more severe at presentation with worse outcomes as compared to upper-middle and high-income countries. Earlier descriptions of CA-AKI in developing regions have come from single center studies and show wide variation in the dominant causes and risk factors. Dehydration, tropical infections, sepsis, exposure to plant and animal toxins and pregnancy related AKI are responsible for the bulf of cases [2,11,12]. A prospective observational study from a large tertiary care center in Puducherry, India identified snakebite and acute pyelonephritis as most common causes of CA-AKI [13]. On the other hand, dehydration, infection and nonsteroidal anti-inflammatory drugs use were recorded as most common precipitating factors in a cohort of CA-AKI from Egypt [14]. In a study from Malawi, 58.8% of patients with CA-AKI had HIV infection, a clinical characteristic that was a surrogate of high prevalence of HIV infection in the community [15]. A review of studies describing AKI in adults in Africa between 2000 and 2015 showed that between 70 and 90% of all AKI cases were community-acquired. The average age of the patients was <40 years, majority presented late with advanced stages of AKI and the mortality ranged between 11.5–43.5% [11]. As compared to those with HA-AKI, patients with CA-AKI in tropical regions are relatively young, frequently male and without significant comorbidities. They frequently work outside in exposed environments and belong to lower socio-economic class with poor access to healthcare. Majority of CA-AKI patients are bread earners for their families. Consequently, the impact of AKI is catastrophic not only for patients but for their families and society as well.

The risk factors and causes of CA-AKI vary with the exposures present in the community, determined largely by the local geographical, social, economic, cultural, and endemic context [16]. The factors can change over time and can affect trends. A large retrospective study on epidemiological trends in CA-AKI between 1990 and 2014 from Pakistan has shown that numbers due to rhabdo-myolysis, malaria, dengue, and obstetric AKI increased over time [17]. Similarly, comparison of epidemiology of CA-AKI between two different time periods (1983–95 and 1996–2008) from Varanasi, India showed increasing trend for AKI due to malaria, sepsis, nephrotoxin exposure, and liver disease related AKI whereas numbers of diarrheal and postabortal AKI decreased [18]. Most of the improvements are related to non-health interventions. For example, postabortal AKI came down and has now virtually disappeared following legalization of abortion in India. Similarly, diarrhea-associated AKI decreased with improvement in supply of potable water. On the other hand, new causes emerge as the ability to diagnose those conditions improve. In a large hospital in Chandigarh, India that was seeing many cases with AKI in the setting of an undifferentiated febrile illness, the availability of a specific and sensitive test allowed the identification of scrub typhus as the cause in a quarter of cases. [19,20] Similar reports emerged from other part of country as well [21]. This has important implications since scrub typhus can be easily treated with cheap and widely available doxycycline. Absent the diagnosis, these patients would have been treated with poly-pharmacy including multiple antimicrobials, which would not only delay recovery but also contribute to other public health problems such as antimicrobial resistance.

One of the largest descriptions of CA-AKI comes from nationwide survey of 2.2 million hospitalized patients in China [22]. The study confirmed the association between epidemiology and outcomes of CA-AKI with socio-economic factors and geo-graphical location. CA-AKI secondary to nephrotoxins or sepsis was more common in Southern areas as compared to more affluent Northern areas. Withdrawal from treatment and access to kidney replacement therapy (KRT) were higher and lower, respectively, in the poorer regions. Overall, 7% of patients with CA-AKI died in hospital and 14% withdrew treatment. Only 64% of those with indication for KRT could receive it. Of those who were discharged, only 40% patients made a full recovery. Estimates based on these data suggest that approximately 1.6 million patients with CA-AKI would have been admitted to hospitals in China in 2013. These data, however, likely underestimate the true burden of CA-AKI as not all patients with CA-AKI would be referred to these hospitals. A proportion might either recover or die in the community or non-traditional primary care settings and hence do not come to medical attention, further underscoring the public health importance of CA-AKI on global scale.

CA-AKI is also encountered in the developed regions of the world, but the clinic-epidemiological profile is closer to that of HA-AKI, unlike the contrast observed in developing countries. Supplementary Table 1, http://links.lww.com/CONH/A41 shows differences in CA-AKI between developing and developed countries. In a recent large retrospective cohort study from Australia, AKI was recorded in 14.8% of patients who were hospitalized for >24 h and had ≥2:2 recorded serum creatinine values [23]. Out of 6477 patients with AKI, HA-AKI was recorded in 55.9%. Patients with HA-AKI were older and had higher comorbidity scores. Underlying chronic kidney disease (CKD) and diabetes were seen with approximately equal frequency (CKD, 24% in CA-AKI versus 23% in HA-AKI and diabetes, 33% in CA-AKI versus 40% in HA-AKI). Meta-analyses comparing CA-AKI with HA-AKI have shown differences in clinical characteristics, length of hospital stay or mortality but are limited by heterogeneity and bias [24,25].

The approach to diagnosis and treatment in the HA-AKI and CA-AKI is different. Hospital-based factors can be identified and addressed in most patients with HA-AKI, as patients are monitored during hospitalization. Strategies based on biomarkers and electronic alerts have been tried to facilitate early identification so that preventive and corrective measures can be implemented in a timely manner. Evaluation and redressal of community-based risk factors for AKI, however, requires a different public health approach that address socio-economic determinants and are integrated in the primary care system.

The Current AKI Definition and Staging Criteria Fail to Adequately Account for CA-AKI

AKI is defined and staged as per Kidney Disease: Improving Global Outcomes (KDIGO) criteria [26]. Quantification of increase in serum creatinine or decrease in urine output over specific time periods is used to define AKI, and the magnitude of these changes or need of KRT are used to stage AKI. Table 1 shows representative components of definition or staging of AKI and their applicability to CA-AKI and HA-AKI. Only a single value of abnormally high serum creatinine is available in most patients with CA-AKI in the developing countries, as patients present with fully established AKI. Documentation of decrease in urine output is also missing. In most of these patients, the diagnosis of AKI is circumstantial and confirmed only later by documenting a decrease in serum creatinine or increase in urine output. This is reflected in observations from the largest CA-AKI cohort from China where the authors showed that almost three quarter of cases of CA-AKI would not have been diagnosed had they applied only the KDIGO criteria for AKI [22]. They used a trend of decreasing serum creatinine after hospitalization to define AKI [22], an approach not endorsed by current guidelines.

Table 1. Components of definition or staging of AKI and their applicability in CA-AKI and HA-AKI settings.

AKI: Representative parameters for defining or staging	HA-AKI	CA-AKIa
Increase in serum creatinine ≥0.3 mg/dl within 48 h	Easily available	Two serial values often not available at admission
Increase in serum creatinine to ≥1.5 times baseline, which is known or presumed to have occurred within the prior 7 days mg/dlb	Easily available	Premorbid baseline values often not available Presumption based on clinical judgement at admission (might be proven wrong later) Decrease in serum creatinine indirectly indicates presence of AKI. However, this is not captured in current definition or staging.
Urine volume <0.5 ml/kg/h for 6 hoursb	Possible in high dependency or intensive care units May not be meticulously possible in general wards. However, can be done when required.	Cannot be objectively ascertained at admission and history often unreliable unless persistent anuria is present
Initiation of kidney replacement therapy (KRT)	Easily documented	Cannot be applied at admission as KRT is not available outside hospitals

^aContext is mainly developing regions where majority of global burden of CA-AKI is present.

^bAlso applicable to AKI staging criteria that use different quantification cut offs.

AKI, acute kidney injury; CA-AKI, community acquired acute kidney injury; HA-AKI, hospital acquired acute kidney injury; KRT, kidney replacement therapy.

A multinational study conducted under the aegis of the ISN 0by25 program [27^{▀ ▀}] evaluated the feasibility of screening patients for AKI (using a symptom-based risk score, point of care creatinine, and urine dipstick) in community health clinics in three low-income countries - Bolivia, Nepal, and Malawi. The true incidence of AKI could not be ascertained due to sheer burden of patients that made screening of all patients impossible. Not surprisingly, these authors were forced to use decrease in serum creatinine to define AKI. Patients were classified at initial evaluation as having acute kidney disease (AKD) if patients did not fit into classification of CKD, AKI or no kidney disease (NKD). Based on decline in serum creatinine over one week, a diagnosis of AKI was made in 47%, 33%, and 25% of patients who were initially classified as AKD, CKD or NKD, respectively. Even though a majority were classed as AKD in this study, there is no widespread recognition or consensus on evidence-based approach to this poorly defined condition. Encouragingly, the study showed feasibility of implementation of point-of-care serum creatinine testing and protocol-based management of high-risk patients in primary care settings. However, actual implementation of these approaches outside research settings, integration with existing healthcare delivery systems in community and impact still need to be ascertained.

Lack of baseline kidney function assessment also confounds subsequent interpretation of recovery. Lack of mechanisms for retrieval of medical records in low-resource settings forbid definitive conclusions. In such situations, the diagnostic dilemma of pre-existing AKD, AKI that has persisted to become AKD or underlying CKD at presentation cannot be addressed at the time of hospitalization [27^{▀ ▀}]. Clinical judgement based on circumstantial evidence, such as presence or absence of ancillary symptoms, signs or abnormal investigations is used to guide diagnostic and therapeutic decisions. The syndromic diagnostic differentials may have to be kept open till serial evaluation of kidney function can be done beyond three months. Even if patients get better and are discharged, ensuring follow up is challenging. In a cohort of 691 patients with CA-AKI but without underlying CKD at a tertiary care hospital in India, 111 (16%) died during hospitalization and 12 (2%) left treatment (Kumar V, unpublished data). Of 568 patients who were discharged from the hospital, about half did not report for follow up, despite repeated attempts to contact them. Of the 291 patients who returned for follow-up, 6 had died and 59 (20.2%) continued to have an estimated GFR <60 ml/min/1.73m² at 4 months. Survivors of AKI are at increased risk of progression to CKD, cardio-vascular events, readmission to hospital, recurrence of AKI and mortality [28,29^{▀ ▀}]. Event rates of 13.9, 19.8, and 3 per 100 person years have been reported for all-cause mortality, CKD, and cardiovascular mortality after AKI [28]. In a propensity matched cohort study from Ontario, Canada, attending AKI follow up clinic was associated with a lower risk of all-cause mortality [30] compared to usual care. Even though AKI as a risk factor for CKD is well known, there are few data about preceding CA-AKI in patients with CKD in low resource settings. In the Indian Chronic Kidney Disease cohort study, about 7% of >4000 participants with CKD enrolled in the study gave a history of previous AKI [31].

In those with very low clinical likelihood of underlying CKD and unavailable baseline values, back estimation of serum creatinine based on estimating glomerular filtration rate equations and assumption of normal GFR (e.g., 75 ml/min/1.73m², as was done in Chinese CA-AKI cohort) has been used to benchmark change in serum creatinine and then, decide on extent of recovery. It is important to note that pre-existing CKD might be missed in such cases and get falsely labelled as incomplete recovery following AKI [32]. These observations make it clear that the current guidelines in AKI need adaptation to document the complete spectrum of CA-AKI and its impact.

Community Acquired AKI is Vastly Under-Researched

Despite being the more common AKI setting, CA-AKI has been studied less well. A PubMed search with title/abstract tag shows just 61 articles for CA-AKI, increasing to 288 on an open search. The current clinical practice guidelines are heavily skewed towards HA-AKI. Of the 75 statements in 2012 KDIGO AKI Guideline, none specifically call out CA-AKI [26]. CA-AKI is mentioned in the sections on introduction, risk assessment and clinical application, and research recommendations for risk assessment advise focus on CA-AKI but fail to delineate any clear approach or plan. The ISN 0by25 initiative did focus on eliminating preventable AKI deaths in low-resource settings but has so far produced few actionable outputs.

Community Acquired AKI: Call for Action

The AKI paradox between high and low resource settings, primarily driven by CA-AKI in tropical and sub-tropical regions, is known but not sufficiently discussed [33]. The 0by25 initiative presented the need to eliminate preventable AKI deaths as a human rights issue. Another approach is to think of CA-AKI as a neglected tropical disease. Better understanding of this condition in all its dimensions and development of sustainable strategies to reduce the global burden and improve outcomes require multidisciplinary collaboration that includes public health specialists, infectious disease experts, policymakers, social scientists, and community members. We cannot achieve global AKI goals like the ones proposed by ISN 0by25 until the research and interventions target risk factors for AKI in the community using an implementation science framework.

The ongoing multiple dynamic interactions between humans and environment on account of changing epidemiology of infectious diseases, emergence of diseases like human coronavirus disease 2019, political or geographical conflicts, war, climate change, natural calamities, scarcity of water etc. have sensitized us of the need to develop resilient healthcare systems. Given that patients with kidney disease are highly susceptible to such threats, nephrologists are well placed to participate in the development of broad-based approaches required to handle such challenges within the context of global health. The field of CA-AKI can set examples in this regard. It should start with qualitative studies in the community and primary care setups. These would help to better understand the factors that determine exposure to community-based risk factors, ascertain barriers to timely healthcare access and uncover infrastructural limitations that impair delivery of care. It is important that the community, healthcare providers and administrative stakeholders build a trustworthy relationship, identify key bottlenecks, and propose novel solutions that improve system-level preparedness to identify and manage AKI. Sustainable systems that seek feedback, support audits and change when needed would be required. Educating the community and healthcare workers will increase awareness and facilitate prevention, early diagnosis and timely referral. Figure 2 and Table 2 show different levels at which care can be offered in the context of AKI. As has been shown by ISN 0by25, the care model needs to change, with greater use of point of care testing and risk stratification using digital tools that can be managed by nonphysician health workers as well as interventions through use of education, training and technology [27^{▀ ▀}].

Figure 2. Acute Kidney Injury.

Levels at which care can be offered, their relation to classification as community or hospital acquired and current guidelines.

Table 2. Possible targets and interventions at various levels of care for Acute Kidney Injury.

Level of care	Target	Intervention
Community	General population Those exposed to environmental, occupational and other external risk factors Pregnant women	Education and awareness Mitigation of exposure
Primary care	Healthcare workers Resources	Education and awareness Identifying high risk patients Prevention Early recognition Emergent management Timely referral Point of care testing
Referral care	Healthcare workers Resources	Geared towards diagnosis and management of complications Increase utilization of low cost therapies like peritoneal dialysis in emergent situations
Follow up care	Healthcare workers Patients	Identifying patients at high risk of incomplete recovery and long term complications Educating patients about need of follow up Mechanisms to ensure regular follow up

AKI, acute kidney injury.

This dissociation between the two forms of AKI, joined by a common definition but different in almost all other ways, speaks of the need for special consideration for CA-AKI, especially in context of low resource settings. The gulf in the magnitude and outcomes on the one hand and the state on research and development, clinical guidance, and care provision on the other between the two forms of AKI is stark. This makes addressing CA-AKI a moral and ethical imperative if the global goals for prevention, early diagnosis and timely management of AKI, and elimination of preventable deaths as articulated in the ISN’s 0by25 initiative are to be achieved. Advocacy is needed for ensuring optimal care and investment in research. The latter needs to span both discovery and implementation research. CA-AKI needs to be given priority by professional societies and organizations involved in the process of developing and setting standards and guidelines. This can only be done by ensuring equitable representation of practitioners from those regions that are inundated with CA-AKI.

Conclusions

The stark epidemiological dissociation between community versus HA-AKI has remained hidden because both are joined together with the same definition. Globally, CA-AKI has not received much attention despite vastly outnumbering HA-AKI in terms of global magnitude and burden of adverse consequences. The current definition of AKI misses out significant proportion of CA-AKI and prevalent guidelines fail to recognize risk factors for CA-AKI. This inequity needs urgent attention. Active collaborative efforts and multi-disciplinary approaches with equitable representation from affected regions are required to find and test Interventions at community level that will improve outcomes in AKI.

Key Points.

• The contextual separation of CA-AKI and HA-AKI is important, as it has implications for prevention, diagnosis and management of AKI.

• Current guidelines align more with HA-AKI and their applicability in CA-AKI settings is poor.

• A multidisciplinary, collaborative approach with participation from community and local healthcare partners is required for meaningful impact.