Chronic kidney disease (CKD) is considered a global public health concern, as approximately 10% of the world’s population is affected by it.¹ With the tremendous socioeconomic burden and insidious nature of this disease, the International Society of Nephrology (ISN), European Renal Association, American Society of Nephrology, and 19 other worldwide experts announced the campaign to add kidney disease as one of the World Health Organization’s (WHO) top causes of death by a non-communicable disease.⁴ In the United States, kidney disease presents in more than 1 in 7 people, which is estimated to be about 35.5 million people.² As one of the most expensive care groups in the Medicare population, reducing the progression of this disease to later stages unlocks savings potential for the healthcare industry and better quality of life for patients affected.³

According to the National Kidney Foundation, CKD is a disease defined by a gradual loss of kidney function over time.⁵ CKD can eventually lead to end-stage kidney disease (ESKD), which often requires renal replacement therapy (RRT), such as dialysis or kidney transplantation. Slowing the progression of this disease assists in reducing complications and presents an opportunity for substantial cost savings.¹ The advancement of CKD can be challenging to detect as the symptoms are often unnoticeable to patients in the early stages. According to the Centers for Disease Control (CDC), 9 in 10 adults with CKD are unaware they have it.² Moreover, in adults with severe CKD, 1 in 3 do not realize they have the condition.² Detecting risk for CKD and its further development requires analyzing data from multiple angles, making it difficult for healthcare providers to sort through and track this amount of data. Predictive analytics offers an opportunity to organize, define, track, and trend data in a way that can deliver real-time proactive insights. Partnering with healthcare providers, the artificial intelligence (AI) of data analytics and machine learning (ML) may provide an avenue for detecting worsening CKD.¹

Several factors contribute to the risks of developing CKD. Genetics, autoimmune disorders, infections, and other conditions can precipitate CKD. In addition, risk factors like diabetes, hypertension, heart disease, heart failure, age, family history, history of acute kidney injury, and use of tobacco products can also lead to the onset of CKD.⁵ Generally speaking, CKD does not develop by a singular reason but is a culmination of social, environmental, and physical factors. Two primary laboratory studies determine the presence of CKD: estimated glomerular filtration rate (eGFR) and urine albumin-creatinine ratio (uACR).⁵ An eGFR test determines how well the kidneys can filter blood to remove waste and distinguishes CKD stages.⁵ Albumin and creatinine are the two substances measured in a uACR test. When kidneys function appropriately, albumin should remain in the blood while the kidneys filter out creatinine. If albumin is leaking into the urine, this could mean the kidneys are damaged.⁵

Once CKD is present, predicting its progression means monitoring a wide array of information. Predictive analytics offer opportunities to organize this data into patterns of predictability to identify those at risk and provide early intervention.¹ Offering treatment options that contribute to these goals assists in delaying the need for RRT and contributes to a better quality of life. Additionally, predictive analytics can aid in creating efficient solutions by pinpointing patients at most significant risk to the appropriate resources such as nephrology, diagnostics, and individualized care strategies to mitigate developing complications or worsening disease.¹ Lastly, AI/ML tools can also find biomarkers and key predictive factors associated with worsening CKD, allowing providers to focus on managing these risks.¹ Predictive analytics can partner with clinicians to provide the individualized care needed in value-based care (VBC), enhance understanding of CKD progression to treat current and future patients effectively, and consistently improve overall outcomes.¹

In developing a treatment plan for CKD, options for care center around slowing the development of the disease. The challenge presents with early identification, as early signs and symptoms of the disease primarily go unnoticed.⁵ A recent systemic review of 13 relevant studies demonstrated that applying AI/ML tools in identifying and treating CKD has shown promise in determining risk estimation, recognizing disease, and providing individualized care.¹ These programs can be interwoven with electronic health records (EHR) and clinical decision support systems (CDSS) to analyze large datasets, including laboratory results, clinical and historical data, and demographics. AI/ML programs can recognize patterns in the data that surpass traditional assessment methods, finding high-risk individuals and allowing for referral for early interventions. Proactive care in patients with CKD can support delaying progression to later stages and provide opportunities for enhanced outcomes. This model aligns with VBC initiatives and cost savings.

VBC initiatives are shifting the healthcare industry from a pay-for-service to a proactive, individualized approach to healthcare delivery. Caring for the CKD population means early identification, intervention, and reducing worsening disease. As CKD progresses, the cost of care increases and the treatments and complications become more significant. According to the United States Renal Data System (USRDS) 2024 Report, Medicare patients with CKD (all ages) accounted for over one-quarter of spending or $95.7B in 2022.³ In examining per person per year (PPPY) spending by CKD stage, costs increase significantly in CKD stages 4-5. PPPY spending is also higher when combined with comorbidities such as heart failure and diabetes mellitus.³ Overall, care costs for this population increase by more than twice what it costs to care for individuals without CKD. Monitoring CKD patients with AI/ML programs can provide efficiency and oversight to highlight patients most at risk–streamlining resource allocation and delivering cost savings to healthcare institutions.⁵

Numerous predictive analytics tools are available in the healthcare sector and are incorporated into EHR systems. An example is the Vasc-Alert software program for patients on hemodialysis. Vasc-Alert is an FDA-approved predictive analytics program that uses data entered in the EHR to determine overall hemodialysis access risk. The data is evaluated to generate a scale of 1 to 10 risk score, providing the clinician with an easily interpretable figure. The higher the score, the greater the risk. Healthcare professionals can facilitate early outpatient interventions for these patients, which helps to reduce complications related to access issues. This approach also enhances the longevity of a patient’s arteriovenous (AV) access. Tools like Vasc-Alert help clinicians analyze data patterns to recognize complications before they worsen.

Utilizing AI/ML tools to identify risk factors for patients with CKD is crucial for healthcare institutions to effectively participate in VBC initiatives and provide the personalized care required. Initiating these practices allows for better quality of life, optimal outcomes, and cost savings. Diagnosing CKD in its early stages is also necessary to begin a treatment plan that slows the progression of the disease and delays the need for costly RRT. In collaboration with clinicians, predictive analytics can lay the foundation for future patient care initiatives. This approach allows for practically identifying, tracking, and trending data, enabling targeted care and therapies. Such strategies are essential for the healthcare industry to remain financially viable in a VBC environment.

References

1. Khalid F, Alsadoun L, Khilji F, Mushtaq M, Eze-Odurukwe A, Mushtaq MM, Ali H, Farman RO, Ali SM, Fatima R, Bokhari SFH. Predicting the Progression of Chronic Kidney Disease: A Systematic Review of Artificial Intelligence and Machine Learning Approaches. Cureus. 2024 May 12;16(5):e60145. doi: 10.7759/cureus.60145. PMID: 38864072; PMCID: PMC11166249.
2. Centers for Disease Control and Prevention. Chronic Kidney Disease in the United States, 2023. Atlanta, GA: US Department of Health and Human Services, Centers for Disease Control and Prevention; 2023.
3. United States Renal Data System 2024 Annual Data Report. National Institute of Health. https://usrds-adr.niddk.nih.gov/2024/chronic-kidney-disease
4. Time for action: recognizing chronic kidney disease as a major noncommunicable disease driver of premature mortality. Wong, Germaine et al. Kidney International, Volume 105, Issue 6, 1144 - 1146
5. (2023, September 11). Chronic Kidney Disease (CKD). National Kidney Foundation. https://www.kidney.org/kidney-topics/chronic-kidney-disease-ckd