|
|
|
Indian Pediatr 2020;57: 276 |
 |
News in Brief |
|
Gouri Rao Passi,
Email: [email protected]
|
Novel Coronavirus (COVID-19) Epidemic
Corona means crown in Latin. The now infamous Corona virus got
its name from a spiky crown of glycoproteins on its surface.
This RNA virus typically infects mammals and birds. It has
caused two previous human outbreaks - SARS CoV and MERS CoV. The
SARS CoV epidemic, which began in a hotel in Hong Kong in
2002-2003, infected 8098 cases with a mortality of 10%. The 2012
MERS CoV outbreak in the Middle East had infected 855 cases with
a mortality of 37%. In the current Coronavirus epidemic in China
the first case was reported in Wuhan on 1st December, 2019 and
as of 9 February, the total number of cases documented
officially had crossed 37,000 and it appears to have a mortality
of 2-3%.
The virus seems to have originated in the Huanan
seafood market in Wuhan in China and was probably transmitted by
pangolins (ant eater) to humans. In the first article published
in the Lancet, the clinical symptoms of 41 patients (median age
49 year) infected with the COVID-19 included fever, cough,
myalgia, hemoptysis and dyspnea. Investigations revealed
lymphopenia, bilateral fluffy shadows or ground glass appearance
on the chest X ray and elevated cytokines and troponin I in the
critically ill patients. Of the 41 patients, one third were
admitted to ICU and 6 died. Diagnosis is being made using RT-PCR
and treatment includes supportive care and empirical use of
antivirals including oseltamivir, lopinavir and ritonavir.
Incubation period appears to be 2-14 days. Unlike SARS and
MERS CoV, which had gastrointestinal symptoms in 20-25%, these
are rare in the COVID-19 infection.However, a subsequent study
published in JAMA analyzed 138 patients of whom 40 were health
care workers. In this series, a patient admitted in a surgical
ward with abdominal symptoms went on to infect 10 more
healthcare workers. The high mortality appears to be due to a
cytokine storm; though, the use of corticosteroids in these
patients did not appear to improve outcomes and also delayed
viral clearance. Medical personal are advised to use fitted N95
respiratory masks to limit exposure besides routine precautions
of hand washing and precautions with aerosol.(The Lancet 29 Jan
2020; JAMA 7 Feb 2020)
The Robot Radiologist
The use of artificial intelligence (AI) in radiology has
recently skyrocketed. A 2018 market survey found that 84% of
radiology clinics in the US were either using or planning to use
artificial intelligence systems. However, the data and
algorithms that these systems use to make a diagnosis are
sometimes inexplicable to humans. This is called the black box
problem. A case pinpoint would be a study published in 2019 in
JAMA Network Open studied 85000 chest X rays in people followed
over 12 years. Raw data is fed into the computer and then the
computer creates its own algorithms to predict outcomes. This is
called deep learning. Impressed by the accuracy of the programs
predictive ability, when researchers analyzed what it was the
computer used to predict mortality, many unusual data was
noticed. For example, one parameter the computer used was
regions below the shoulder which has no known medical
significance. Retrospectively it is felt that the parameter
represents flexibility and hence may predict mortality. This
discordance of human and computer-aided thought process is
called the black box problem since it is virtually invisible to
human understanding.
What is still unclear is that if
medical AI systems make a mistake who will bear the
responsibility. One way around the problem is to develop
transparent systems which explain the factors taken into
building the algorithm at every stage. Another variable in
medical AI systems is that they change and improve over time as
they get access to more data and their performance is in
constant flux. For now the FDA has developed guidelines for
algorithms which evolve over time.
It appears that AI may not
replace radiologists in the near future, but “radiologists who
use AI will replace radiologists who do not.” (Scientific
American 1 February 2020)
Drone Delivery of Blood
Products
In the East African nation of Rwanda,
medical history is being quietly written. On 21 December, 2016,
a 2-year-old girl with severe malaria became the first person to
receive a drone delivery of packed RBCs. Since then, more than
4000 units of blood products have been delivered using dronesby
a US-based startup called Zipline. This technology has reduced
the time to deliver blood products in remote areas from nearly 3
or more hours to barely 15 minutes. Till some years ago maternal
mortality due to postpartum hemorrhage was a huge problem in
Rwanda. The rapid delivery of blood has helped to save precious
lives.
It all began when robotics expert Keller Rinaudo
and aviation expert Will Hetzler met public health researcher
Zachary Mtema in Tanzania. For many of the critical medical
problems like availability of blood products, anti-snake venom
and anti-rabies immunoglobulins, drone delivery seemed to be an
ideal solution. A company was founded and a deal struck with the
Rwandan government to build a distribution center near Muhanga.
The companies drones, deliver medical supplies within an 80 Km
radius of a distribution center. The cost per service is same as
the previous motorcycle service but more reliable. In India,
Maharashtra has announced that Zipline will provide emergency
medicines in the entire state this year. These drones have been
listed on Time magazine’s best inventions of 2018.
(www.who.int 12 June 2019)
|
|
|
 |
|
