Easily accessible AEDs can save lives
We’ve learned a lot about COVID-19 over the last three years, but many mysteries remain. For example, long COVID is common and difficult to diagnose.
In the National Center for Health Statistics pulse survey from January 4 – 16, 2023:
Researchers published a major review of long COVID in Nature in January 2023. They found that long COVID affects at least 10% of those who are infected and identified 200 long COVID symptoms that impact organ systems. They also found that the COVID-19 virus can cause cardiovascular events (heart attacks and strokes), blood clots, chronic fatigue syndrome and autonomic nervous problems such as a high heart rate and low blood pressure when standing up. Some sufferers have a significant decrease in cognitive function.
There are several potential causes for long COVID, and some patients might have more than one cause. These include:
Those at highest risk of developing long COVID include women, and Latinos, and people with the following conditions:
Nonetheless, a third of those with long COVID have no known predisposition.
Reinfection increases the risk of long COVID, even in those who are vaccinated. While those who were critically ill with COVID-19 are more likely to have persistent symptoms, many with long COVID had mild cases.
Kids get long COVID too, but because they are more likely to have negative PCR tests, they may not meet some of the criteria for diagnosis. This review points out that for many, the diagnosis of long COVID will persist.
Some people with long COVID can be diagnosed, but many will not have a positive test result to “prove” the diagnosis. Researchers also note that some people with symptoms of long COVID are misdiagnosed with a mental health disorder. The good news is that some people will respond to medical therapy for symptoms, and researchers are studying the antiviral Paxlovid as a potential treatment for long COVID. Paxlovid decreases the risk of long COVID by about a quarter in both vaccinated and unvaccinated patients.
A BMJ study of medical records in Israel, shows that “patients with mild COVID-19 are at risk for a small number of adverse health outcomes, most of which are resolved within a year from diagnosis.”
The country was transfixed last month when Buffalo Bills safety Damar Hamlin suffered a cardiac arrest after a tackle in a Monday night football game with the Cincinnati Bengals. He likely had commotio cordis — an external blow to the chest at the exact instant of the heart rhythm, which can lead to a fatal abnormal heart rhythm. On-field medical personnel immediately began cardiopulmonary resuscitation (CPR), and he was given a shock using an automated external defibrillator (AED), which restored his normal cardiac rhythm. He was released from the hospital nine days later and his prognosis is good.
Fatal heart rhythms or sudden cardiac death are commonly caused by a heart attack, and a rapid response makes survival much more likely. There are 350,000 cardiac arrests annually in the U.S. About 40% of those suffering from a cardiac arrest receive CPR, whereas only about 10% are treated with an AED. Many more lives could be saved if more people were treated with AEDs, which require no medical training.
The responder applies two adhesive leads to the chest, and the machine detects existing rhythm and gives clear instructions about what to do. If the person has a heart rhythm and wouldn’t benefit from a shock, the machine will not administer a shock.
Current resuscitation guidelines when someone has lost consciousness and has no pulse are:
Guidelines no longer prioritize mouth-to-mouth resuscitation.
All states have Good Samaritan laws to protect responders from lawsuits based on their efforts to help resuscitate people in distress. Some states, including California, require AEDs in many public buildings.
Machine learning and artificial intelligence (AI) will change medical care. Over the past two decades, there has been a significant increase in the medical devices approved by the Food and Drug Administration (FDA) that incorporate AI and machine learning.
A preeminent economist and three consultants published a report last month suggesting that use of AI could decrease medical costs by 5% – 10%, using technology already available and implementable in the next five years. The report states that these savings would be about 40% from administrative savings (9% – 19% of administrative costs) and 60% from reducing medical costs (4% – 8% of medical costs). Some examples of AI use in healthcare:
The report does not include the initial capital cost of incorporating AI in medical and administrative processes in medical care. Indeed, the adoption of AI could increase total healthcare costs over the coming years.
AI systems are trained on real world data, which incorporates substantial structural racism. So we will need to guard against inadvertently increasing existing disparities.
Finally, I’m worried that some of these administrative savings will be lost as we initiate an ever-escalating AI battle. For instance, insurers can use AI to automate prior authorization — but what if providers implement AI too? Will “battling bots” actually lower healthcare costs?
One study published this month shows that the hassle factor of prior authorization lowers medical costs, so automating prior authorization could paradoxically raise costs. On the other hand, the use of Chatbots to increase touchpoints resulted in an increase in childhood vaccination in a poor neighborhood in Chicago and will likely result in less childhood illness and reduce healthcare demand.
Earlier promises of AI revolutionizing healthcare like IBM Watson would revolutionize oncology care proved premature, and providers are likely to move slowly and require substantial validation before implementing AI and machine learning in clinical practice.
I suspect that we will see the incorporation of AI and machine learning to decrease waste in healthcare, but adoption will be slower than the report projects and the savings could be more beneficial to providers or health insurance plans than the employers that sponsor health insurance plans.
Researchers at Clarify Health, a software and analytics company, reported last week that their review of claims data shows that orthopedic surgeons who performed more than 100 hip or knee replacements on those insured by commercial payers from 2017 to 2020 had dramatically better outcomes and lower costs in 2021 than low-volume orthopedic surgeons, who performed fewer than 10 procedures. The sample was over 66,000 hip replacements and 110,000 knee replacements, about one in seven procedures performed in the U.S. The majority of hip and knee replacements are done on those over age 65 and are covered by Medicare, and these were not included in this study. The researchers standardized costs and controlled for other medical conditions.
Employers should promote transparency initiatives to inform patients about surgical volume when they are deciding which surgeon to consult.
Jeff is an internal medicine physician and has led WTW’s clinical response to COVID-19 and other health-related topics. He has served in leadership roles in provider organizations and a health plan and is an Assistant Professor at Harvard Chan School of Public Health.
