Machine learning is a type of artificial intelligence that helps computers learn from data and get better over time without being programmed directly. Recently, machine learning has started giving useful solutions in healthcare, especially in handling complex operations and public health problems.
Hamsa Bastani, a professor at the Wharton School, has worked on new machine learning methods for healthcare operations and public health decisions. Her research focuses on using data to improve healthcare supply chains, clinical trials, and resource management during emergencies like the COVID-19 pandemic.
One example is Bastani’s work on improving drug supply chains in poorer countries like Sierra Leone. Even with limited data, machine learning helped predict drug demand more accurately. This made sure that medicines were shared fairly. The U.S. faces different challenges, but the idea is the same: machine learning helps predict needs better and manage resources well, reducing waste and shortages.
Machine learning is also changing how clinical trials are run and how treatments are adjusted for each patient. Bastani’s research includes a clinical trial design that uses both early results and final outcomes to make better decisions during trials. This method has cut trial costs by about 16% without making more mistakes, which helps bring new drugs to patients faster while keeping research accurate.
For doctors and health managers in the U.S., these changes mean faster medical advances and treatments better suited to each person. By predicting how patients will respond to treatments, machine learning can lower hospital readmissions and improve overall health results.
The COVID-19 pandemic showed how machine learning can help with public health efforts. Bastani worked with the Greek government to create “Eva,” the first national system using reinforcement learning for COVID-19 testing. This system used real-time data to decide where to send tests, making testing more efficient and helping leaders make better choices during uncertain times.
This system was transparent and supervised by humans, which helped build trust among health officials. In the U.S., similar AI tools could be used to handle future health emergencies, improve vaccination plans, or manage hospital resources during busy periods.
Pharmaceutical supply chains and managing medical inventory are complicated and can be disrupted, as seen during the early COVID-19 outbreak. Machine learning can help improve demand forecasting and inventory control in these areas.
Based on Bastani’s work in countries with limited data, U.S. health systems can use transfer learning. This technique adjusts existing prediction models for new data, even when data is scarce or messy. It helps predict drug use, equipment needs, and staffing more accurately.
Using machine learning to forecast and distribute supplies allows U.S. hospitals to cut costs, avoid shortages, and keep operations running smoothly in both normal and emergency situations.
Machine learning also helps improve how patients stay involved in their care by giving personalized suggestions. Usual algorithms sometimes suggest too many new or unrelated options, which can make patients lose interest.
Bastani developed algorithms that limit how much the system tries new options, keeping patients interested without overwhelming them. This method can be helpful in digital health services, telemedicine, or patient portals, where keeping people engaged matters for following treatments and tracking health.
Many studies in the U.S. show that AI helps in early diagnosis, predictions, risk assessments, and treatment planning. A study by Mohamed Khalifa and Mona Albadawy looked at 74 studies and found eight key clinical prediction areas where AI is useful. These include early diagnosis, forecasting disease progress, and predicting treatment results and risks like readmission or complications.
Fields like oncology and radiology, which use lots of images and data, have especially benefited. AI can analyze medical images accurately and quickly, leading to earlier and more precise detections. This helps doctors plan better treatments and improves patient health.
For healthcare managers, using AI here can mean smoother workflows, happier patients, and safer care.
Using AI in healthcare has challenges. Problems like data quality, bias in algorithms, understanding how AI works, and following rules must be carefully handled. Researchers and leaders stress the need for openness, human control, and ethics to make sure AI tools are safe and reliable.
In mental health, AI can help with virtual therapy and early detection but raises concerns about privacy and keeping human care and empathy in treatment. These issues also matter in physical health care.
Strong regulations and ongoing checks are needed to use AI safely and keep public confidence, especially in the U.S. where healthcare rules are strict and patient safety is a priority.
Machine learning and AI also help by automating office and administrative tasks in healthcare. This lowers the workload on staff so they can spend more time caring for patients. This helps operations run better and can improve patient satisfaction.
For example, Simbo AI offers AI-based phone systems for medical offices. These smart phones handle appointment booking, patient questions, and message routing without human help. For busy offices in the U.S., this reduces wait times and missed calls, improving communication and patient retention.
Besides phones, AI can manage scheduling, records, billing questions, and insurance checks. It can predict when patients might miss or cancel appointments, helping staff manage calendars more efficiently.
AI tools that work with electronic health records can also automate data entry and coding, lowering errors and freeing staff from paperwork. This is important in the U.S. where billing and documentation are complex.
Overall, AI-driven workflow automation helps save money and creates better experiences for patients in American healthcare.
Healthcare often struggles with having the right number of staff and necessary resources. AI models that use past data and current information can predict how many patients will need care and how severe their conditions might be. This helps managers schedule staff in a smart way.
In U.S. emergency rooms and clinics, AI staffing tools can cut overtime costs and patient waiting times. Reinforcement learning models, like the one used in Greece for COVID-19 testing, can change resource distribution based on patient needs in real time.
For clinic owners and managers, these predictions improve how operations respond to changes, lower burnout among healthcare workers, and support more consistent care.
Using machine learning in healthcare management shows promise for better decisions and more effective operations in the United States. As data grows bigger and more complex, these algorithms will be needed to find patterns, predict results, and improve daily work.
For medical office managers and IT staff, investing in AI products like Simbo AI’s phone systems or clinical support tools can help lower costs and boost patient involvement. But success needs careful planning, regular review, and training to match AI tools with goals and regulations.
Creating a work culture based on data and ethical AI use will improve healthcare quality and make it last longer.
In short, new machine learning methods are being used in many parts of U.S. healthcare. These include supply chain management, clinical trials, diagnostics, public health responses, workflow automation, and patient engagement. With more research and careful use, machine learning can help make healthcare more effective, efficient, and focused on patients.
Hamsa Bastani’s research primarily focuses on developing novel machine learning algorithms for data-driven decision-making, with applications to healthcare operations, social good, and revenue management.
Machine learning can improve healthcare supply chain performance by enhancing demand forecasting, which is crucial for effective management in pharmaceutical supply chains, especially in low- and middle-income countries.
Challenges include limited data availability, ensuring equitable distribution, and the need for innovative machine learning techniques to address these issues effectively.
The proposed design combines data from surrogate and true outcomes in clinical trials, potentially reducing costs by 16% while maintaining accurate error rates, thereby improving drug approval decision-making.
Reinforcement learning is utilized to dynamically allocate tests within a national COVID-19 testing system, optimizing resource use under non-stationary conditions while fostering transparency and trust among stakeholders.
Bastani’s work applies machine learning models to various public health issues, including supply chain optimization, effective patient interventions, and managing resources in healthcare systems.
Bastani has received numerous awards including the Wagner Prize for Excellence in Operations Research Practice and the Pierskalla Award for Best Paper in Healthcare, multiple times.
Understanding customer disengagement is crucial for online platforms, as poor recommendations can lead to abandonment; Bastani proposed algorithms to enhance engagement while recommending products.
Bastani employs machine learning to analyze deep web data, revealing trafficking patterns in commercial sex supply chains, which assists law enforcement and policymakers in targeting interventions.
The course aims to improve understanding of AI’s role in business transformation, discussing its applications and ethical governance frameworks, catering to students without a technical background.
Simbo is using AI agents to automate all phone related workflows. Connect now to know more.
Schedule AI Agents Demo Now© Since 2019, Simbo AI.
