The topic of healthcare costs is too big to be addressed by one person, leave alone in any one blog-post.
However, there are some areas which directly lead to high costs and I would like to describe three of those in this post.
Incorrect or incomplete diagnosis:
When a diagnosis is not done properly, the patient will not be treated for the right problem, and consequently will not have the right outcome.
Experienced physicians carry innumerable differential diagnoses in their minds (through years spent seeing multiple variations of the same complaint) and can arrive at the right diagnosis fairly quickly.
But what about junior physicians, nurse practitioners and the like?
With artificial intelligence, we can bundle all the learnings of experienced physicians and provide a virtual assistance to the junior physician. This will not only help the physician not miss a diagnosis, but know the right questions to ask and present a confident diagnosis.
The urgent care centers – growing at more than 50% in the last 5 years – generally do the diagnostic part and are staffed with junior physicians. We feel the AI-driven diagnoses is especially valuable at such centers.
Not ordering the right tests:
An extension of the diagnosis is the intelligence that can prescribe the tests to be ordered for a specific condition. When the virtual assistant lists out the tests to be ordered, the physician has a ready check-list and need not ask for “just in case” tests.
The number and nature of tests ordered has to be just right. If a required test is not ordered, the data-driven diagnosis and subsequently the treatment will be off-target. Sure, it will help keep the costs lower initially, but the incomplete treatment (or treating an un-present condition due to lack of information) will result in re-admission of the patient and higher costs.
Ordering additional, unnecessary tests will directly increase costs to the patient.
An AI-driven virtual assistant will help the physician get the tests just right, and keep the costs low.
High costs of drugs:
Another extension of the virtual assistance would be the prescription part. For each diagnosis the assistant can list out the treatments to be prescribed. It is simple functionality that can be built into a software solution to connect to the hospital drug formulary. Thus, the physician can see the projected costs of drugs, and lab tests even as they are in prescription mode.
Currently doctors (and surgeons) are generally not aware of the costs. Their single-minded focus is on getting the best treatment for the patient so as to ensure the best possible outcome, regardless of cost.
A simple re-ordering functionality in the virtual assistant can rank the drugs (to be prescribed) in the order preferred by the patient. If the patient has consented to the use of a similar generic drug, then the system can rank a lower-cost generic equivalent of a branded drug higher up the list.
Antibiotic prescriptions:
Overall 1 in 7 patients receive unnecessary antibiotics (source). In a 2016 US survey, according to the same source, 23% of antibiotic prescriptions were not medically justified, only 36% were potentially appropriate and 28% were not associated with any documented diagnosis.
Antibiotic resistance leads to poorer patient outcomes and wrong prescription leads to extended hospital stays, all leading to increased healthcare costs.
Artificial Intelligence can help in bio-surveillance, insights on infections and the right prescriptions.
Extended hospital stays:
Extended hospital stay is directly related to higher healthcare costs. Whether due to incomplete diagnosis, incorrect prescription, inaccurate tests ordered or other, an intelligent virtual assistant can help with data-driven, expert guidance to the physician or nurse practitioner and help keep costs down and improve patient outcomes.
Recent Comments