As is well known in the industry, the hospital discharge process includes many components that could impact the success of a patient’s recovery well after they leave the hospital. In some cases highly complex factors may prolong a patient's stay in the hospital and these factors can be difficult to identify early in the hospital course. Though case managers have a remarkable understanding of their patient’s needs, they are often burdened with large patient loads and discontinuous scheduling. Case managers and other professionals involved with care coordination may therefore benefit from systems that identify and address these complex and often subtle factors earlier in their hospitalization.
For example, imagine a new patient who does not recognize or admit that they can care for themselves. The care team themselves may not understand the patient’s awareness until well into the hospitalization, as it might take several conversations and processing of the current clinical predicament. Separately, if the patient also lacked close family support, the discharge suddenly becomes very complex. Each of these subtle situational variables must be uncovered and factored into the discharge plan as early as possible. Systems that can recognize, predict and update these interlocking barriers in all their nuance and subtlety can help provide more time for providers to identify interventions and reduce stress for front-line care managers.
Uncovering complex insights
Many of the clues needed to help make these complex judgements do actually exist within the vast pool of written documents, often referred to as “unstructured data", in the electronic health record (EHR). However, finding this information can be significant work for individuals involved with care coordination. Parts of the information may be hidden in a physician’s progress note, or in the written reason for a consult. It could be located in a nursing flow sheet or a case manager referral. The amount of data available, and where to find it, can be overwhelming and difficult to quickly analyze within the frenetic clinical environment.
Artificial Intelligence (AI) is the ability of a computer to perform tasks typically requiring human intelligence. Natural language processing (NLP), a sub-field of AI, gives machines the ability to read and understand the human language. AI helps solve this problem by being able to read and process everything that has ever been written, said, or documented about a patient in order to garner insights and provide a coherent explanation about what might be occurring. NLP can be used to read clinical notes and bring structure to previously unstructured data. Increasingly these technologies are able to interpret subtle clinical discharge, such as the example above, through a combination of direct and indirect reflections in the EHR, as well as information gathered from patient engagement tools, like chatbots, and external social determinants of health data.
AI reveals SDOH of patients
As research shows, 70% of health outcomes are determined by health related social needs-- factors truly outside of hospital walls. Identifying the social determinants of health (SDOH) while the patient is in the hospital enables greater opportunities to coordinate with services in the community. AI systems can read clinical notes to identify many SDOH, and then assist with locating the best community based organizations.
Reducing length of stay
Often there may be numerous barriers earlier in a hospitalization, and hospitals can reduce the barriers. Optimizing the hospital discharge process is important for many reasons. A key benefit of reducing hospital discharge barriers is reducing the length of stay for a patient. This results in reduced costs for hospitals and increased efficiency. On average in the US, it costs a hospital $2,260 for each inpatient day, so reducing the number of unnecessary days a patient spends in the hospital can result in large savings for health systems.
How Pieces supports Case Managers
The use of technologies like artificial intelligence and natural language processing is increasingly important to health systems. Solving for issues like how to improve the hospital discharge process can seem overwhelming, but can be tackled with the right partner. Pieces Predict is a cloud-based software that uses AI and NLP to extract actionable data from the electronic medical record (EHR) in order to reduce the time doctors, nurses and case managers need to “hint down” information in the chart. This allows them to work more efficiently and provides them with predictions that can improve patient care and timely hospital discharges.