Medically Speakin

HOW AI PROGRAM BOOSTS PATIENT’S SATISFACTION

Published

on

The healthcare business is in desperate need of transformation. There are various ways in which technology may improve clinical workflows, improve experiences, and elevate overall care, from chronic conditions like heart disease and diabetes to risk assessment.

Artificial intelligence is developing as a plausible answer to such challenges, as physicians nowadays experience high rates of burnout and people demand more attention from their doctors. It frees up doctors’ time, allowing them to provide care with greater attention and efficiency.

Predicting senior falls, altering oncology, increasing access to care in distant places, and much more are all examples of how the subject, which includes machine learning and deep learning, is already making ripples in the world of healthcare today.

Diagnosis is the First Step in the Treatment Process

When it comes to medical diagnoses, a variety of algorithms have shown promise. These include Machine Learning (ML) methods, such as neural networks or clustering algorithms, in which a computer is trained on a huge data set. The data could be X-rays with the words “illness present” or “disease not present” pre-labeled by an expert. The computer updates the settings of its algorithm with each iteration. After the algorithm has been trained, it is validated on a portion of the dataset that was kept back throughout the training process.

When it comes to tasks that demand more general knowledge, humans will always beat machine learning. Giving medical practitioners more tools is the goal of AI diagnosis.

Not only will providers have access to these tools, but when the machine malfunctions, a doctor will be able to provide input that can be used to improve the system’s accuracy. The algorithm will profit from the collective wisdom of the medical community.

Natural Language Processing in Electronic Medical Records

Most people recognise the potential of an electronic medical record (EMR), yet doctors now spend more than twice as much time inputting data as they do talking with patients. Natural Language Processing (NLP) is a branch of AI that is intended to assist in this endeavour.

NLP seeks to derive meaning from spoken or written language in order to offer meaningful information to a computer system. Rule-based approaches were utilised in earlier attempts at NLP, whereas modern NLP algorithms incorporate Machine Learning.

3) Scheduling of Nurses

The majority of nursing units schedule nurses by hand. Creating a timetable that works for you might be a difficult undertaking. Overstaffing is costly, whereas understaffing causes safety concerns and burnout. Hospitals frequently compensate by adding agency nurses to their roster, which is even more costly.

An algorithmic method could optimise the schedule by not only arranging the right nurses at the right times, but also fairly and transparently accounting for the nurses’ preferences. As a result, there is better safety at a lower cost, and the management has more free time.

4) Patient Surveys for Sentiment Analysis

To gauge patient happiness, most healthcare institutions use a patient satisfaction survey. A numerical score is useful, but what if we could explain a patient’s feelings in a more nuanced way? Gratitude, rage, contentment, and frustration could be measured instead of “strongly agree” to “strongly disagree.”

Sentiment analysis assigns an emotion to unstructured material, such as survey responses, using Natural Language Processing and Machine Learning. Reviewing patient survey responses and quickly exposing issues that need to be addressed is one apparent application. Aggregating and monitoring how this data evolves over time allows you to understand the impact of process improvements on a larger scale.

5) Robotic Nurse Assistant

You might not want a robot to take the position of a nurse, but what if a robot could take care of some of the tedious and time-consuming tasks, allowing the nurse to spend more time with their patients?

Moxi, a robot built by Diligent Robotics to transport supplies and samples, has proven to be highly popular. The robot isn’t meant to be scary, and it goes about its business quietly and effectively.

6) Automated Pharmacy

Patient safety is jeopardised by medication errors. A pharmacy robot may not be as human as Moxi, but it is surely willing to operate continuously and with extreme precision.

UC San Francisco has established a robotic pharmacy that dispenses medications and delivers them to nurses’ stations using robots with a claimed accuracy of 100 percent. Prescriptions are received, filled, and delivered by the machine using autonomous robot “tugs.”

This not only improves patient safety, but it also transforms the pharmacist’s function. More time spent advising patients and talking with doctors regarding patient care implies less time spent filling orders.

7) Financial Interactions

Financial interactions account for a significant portion of the patient experience. Improving this is a fantastic way to enhance the patient experience.  Patients demand transparency and simplicity, as well as flexibility in some circumstances. Calculating the cost of an operation may be simple for an algorithm with sufficient previous data to evaluate, but when government rules and insurance contracts are included in, even R2-D2 would give up.

So, where shall we begin? Insurance companies are starting to employ AI to seek for indications of fraud in the financial sector. Although it may be a stretch to call this a patient happiness issue, fraud has been related to patient damage.

Patient satisfaction is an extremely important metric for hospitals. For improving the quality of patient-centered health care the key drivers behind patient satisfaction is a critical initiator. Patients appear to want speedy and tailored medical care, and AI systems may be able to meet those needs. As proven by the Penn State researchers’ machine learning model, AI has the ability to speed up the process of addressing and enhancing patient satisfaction.

Presently, AI technology is in the middle of a “hype cycle” — a period of over-inflated claims which will eventually give way to a “trough of disillusionment.” Hopefully, this article leaves you with a realistic picture of the tools available and the actual, incremental changes that can be made. Whether or not those changes improve the doctor-patient relationship, and in turn, the patient experience and satisfaction.

The writer is the Co-Founder & CEO, Shubu.ai.

The Daily Guardian is now on Telegram. Click here to join our channel (@thedailyguardian) and stay updated with the latest headlines.

For the latest news Download The Daily Guardian App.

Pain after knee replacement surgery: When to see a doctor?

Published

on

Total Knee replacement surgery is now widely accepted as a precise, repeatable and successful surgery. A large number of cases of knee replacement surgery are carried out in India itself. With better trained doctors and good hospital infrastructure, the results of knee replacement surgery are getting better every day. However, there are a few patients who are not satisfied with their knee replacement surgery and they continue to experience pain. We need to identify the reasons for this and ensure that this does not happen.

Q. What to expect after surgery?

A. With better tissue handling techniques and accelerated rehabilitation methods, most patients can start mobilising 3-5 hours after surgery. There is some pain of the incision and sutures. Post surgery as the patient starts exercising, the pain gradually becomes much lesser and depending upon the person’s body type, age and deformity before surgery, it should go away within 25 to 30 days. With robotic surgery, the mobilisation is much faster and the amount of pain is much lesser and patients bend their knees much faster and more. Some might take a little longer than the others, especially if their muscles and bones are weak or they are elderly. Even comorbidities like diabetes or kidney disease can slow your recovery. If a person doesn’t have the above-mentioned conditions, then they should not have any pain and the joint should start feeling normal.

Q. What to do after discharge?

A. Here are the few tips a patient undergoing knee replacement surgery should follow to speed up his/her recovery process:

Follow doctor’s instructions diligently:  The patient should make sure that he/she is adhering to the doctor’s instructions. Many times people often ignore their surgeon’s suggestion and become doctors on their own by self-medicating. Such instances often yield poor results.

Exercising properly: Most patients can begin their knee exercise within hours of the surgery. But, they have to ensure that they are not overdoing the exercise because even that can cause some stretch pain or can hamper the alignment.

Avoiding falls: This is mostly experienced by elderly people. They should use crutches, a cane, a walker, and handrails to get around postoperatively until their strength, balance, and flexibility have improved.

Q. When to see a doctor?

A. The big question is of persistent pain even after knee replacement surgery. Let’s understand the causes for pain after surgery also. In the initial period, the pain is of the incision and the surgical invasion itself. As that goes down gradually with exercising, the pain can be coming from the spine or the muscles which are weak. Pain after that could be due to weak muscles – mainly the quadriceps of the thigh muscle which brings the leg straight up.

One very serious complication can be infection, which of course in any good centre now with the operation theatre discipline and bacteria free filtered air in OTs should not happen but if a patient has any redness with pain or swelling, high fever then he/she must immediately consult the doctor. Infections require immediate treatment to prevent further serious complications. Although the rates of infection in any good center are less than 1 percent in world, there is no harm in consulting the doctor once and ruling out these conditions.

Another reason for dull aching pain that goes on for months together is malaligned component or a knee which is not balanced properly. With the use of robotics, the artificial components put inside are actually balanced throughout the range of movement and the stretch on the ligaments is accurately assessed on the operation theatre itself. This ensures correct tracking of the knee just like the natural knee joint and hence no pain.

The writer is Orthopaedic & Robotic Joint Replacement Surgeon, NHS Hospital, Jalandhar

Continue Reading

STUDY SHOWS HOW MEDITATION IMPROVES BRAIN FUNCTIONING

Published

on

Just a few weeks of meditation can improve human brain functioning, according to a recent study led by scientists at Binghamton University and the State University of New York.

The findings of the study were published in the journal ‘Scientific Reports’. Millions of people around the world seek mental clarity through meditation, most of the following or inspired by the centuries-old practices of Buddhism.

Anecdotally, those who meditate say it helps to calm their minds, recenter their thoughts, and cut through the “noise” to show what really matters.

Scientifically, though, showing the effects of meditation on the human brain has proved to be tricky.

The study from Binghamton University’s Thomas J. Watson College of Engineering and Applied Science tracked how practising meditation for just a couple of months changed the brain patterns of 10 students in the University’s Scholars Program.

The seed for the research came from a casual chat between Assistant Professor Weiying Dai and lecturer George Weinschenk, MA ‘01, PhD ‘07, both from the Department of Computer Science. Weinschenk is a longtime meditation practitioner whose wife worked as an administrator at the Namgyal Monastery in Ithaca, which is the North American seat of the Dalai Lama’s personal monastery.

“I developed very close friendships with several of the monks. We would hang out together, and I even received instruction from some of the Dalai Lama’s teachers. I took classes there, I read a lot and I earned a three-year certificate in Buddhist studies,” he said.

Continue Reading

BLOOD-PRESSURE DRUGS COULD IMPROVE COLORECTAL CANCER: STUDY

Published

on

A new study has suggested that common blood-pressure drugs may improve survival for patients with colorectal cancer. The findings of the study were published in the journal ‘Cancer Medicine’. After reviewing outcomes of almost 14,000 patients with colorectal cancer, researchers determined that ACE inhibitors, beta-blockers and thiazide diuretics were all associated with decreased mortality.

They also found that patients who took their blood-pressure drugs consistently were less likely to die from their cancer. The researchers emphasised that more research is needed to validate the connection between blood-pressure drugs and better outcomes.

But they said they are hopeful that the drugs could offer a new, low-cost way to improve care for patients with stage I-III colorectal cancer. “Cost-effective solutions to prolong cancer survivorship in older patients may lie in commonly used medications,” said researcher Rajesh Balkrishnan, PhD, of the University of Virginia School of Medicine’s Department of Public Health Sciences. “However, we need further confirmation of these findings through clinical trials,” added Balkrishnan. Colorectal cancer is the third most commonly diagnosed cancer in the United States. The American Cancer Society estimated that this year there will be 104,270 new cases of colon cancer and 45,230 cases of rectal cancer in the US.

A rise in the number of younger people developing colorectal cancer recently prompted the US Preventive Services Task Force to reduce the recommended age for first screening for the disease to 45 from 50. (Colon cancer killed Chadwick Boseman last year.)

High blood pressure is common among patients with colorectal cancer, but there has been little research into the potential effect of blood-pressure drugs on patients’ outcomes.

Researchers at UVA Cancer Center and the Universidade de Sao Paulo Instituto do Cancer do Estado de Sao Paulo wanted to change that, so they conducted a large retrospective analysis.

They used the Surveillance, Epidemiology, and End-Results (SEER) Medicare database to review outcomes of 13,982 patients ages 65 and older who were diagnosed with colorectal cancer between January 1, 2007, and December 31, 2012.

The researchers found that ACE inhibitors and thiazide diuretics appeared to provide the most significant benefit to patient survival and outcomes, while there did not appear to be similar benefits from calcium-channel blockers.Patients’ adherence to their blood-pressure regimen also appeared important: “Our results show an association between increased adherence to [blood-pressure] medications and reduced … mortality in patients starting these medications after stage I, II or III CRC diagnosis relative to those who did not,” the researchers wrote.

Continue Reading

Findings help explain why vaccinated people are at low risk during Delta surge

Published

on

Despite causing a surge in infections this summer that has resulted in thousands of hospitalisations and deaths, the delta variant of the virus that causes Covid-19 is not particularly good at evading antibodies generated by vaccination, according to a study by researchers at Washington University School of Medicine in St. Louis.

The findings, published in the journal Immunity, help explain why vaccinated people have largely escaped the worst of the delta surge. The researchers analysed a panel of antibodies generated by people in response to the Pfizer Covid-19 vaccine and found that delta was unable to evade all but one of the antibodies they tested. Other variants of concern, such as beta, avoided recognition and neutralisation by several of the antibodies.

In previous studies, co-senior author Ali Ellebedy, PhD, an associate professor of pathology & immunology, of medicine, and of molecular microbiology, had shown that both natural infection and vaccination elicit lasting antibody production. But the length of the antibody response is only one aspect of protection. The breadth matters, too.

An ideal antibody response includes a diverse set of antibodies with the flexibility to recognize many slightly different variants of the virus. Breadth confers resilience. Even if a few antibodies lose the ability to recognize a new variant, other antibodies in the arsenal should remain capable of neutralizing it.

“The fact that delta has outcompeted other variants does not mean that it’s more resistant to our antibodies compared to other variants,” said co-senior author Jacco Boon, PhD, an associate professor of medicine, of molecular microbiology and of pathology & immunology. “The ability of a variant to spread is the sum of many factors. Resistance to antibodies is just one factor. Another one is how well the variant replicates. A variant that replicates better is likely to spread faster, independent of its ability to evade our immune response. So delta is surging, yes, but there’s no evidence that it is better at overcoming vaccine-induced immunity compared to other variants.”

To assess the breadth of the antibody response to SARS-CoV-2, the virus that causes COVID-19, Ellebedy and colleagues- including co-first authors Aaron Schmitz, PhD, a research specialist; Jackson S. Turner, PhD, an instructor in pathology & immunology; and Zhuoming Liu, PhD, a staff scientist- extracted antibody-producing cells from three people who had received the Pfizer vaccine. They grew the cells in the laboratory and obtained from them a set of 13 antibodies that target the original strain that began circulating last year.The researchers tested the antibodies against four variants of concern: Alpha, Beta, Gamma, and Delta. Twelve of the 13 recognised Alpha and Delta, eight recognised all four variants, and one failed to recognise any of the four variants.

Scientists gauge an antibody’s usefulness by its ability to block viruses from infecting and killing cells in a dish. So-called neutralizing antibodies that prevent infection are thought to be more powerful than antibodies that recognize the virus but can’t block infection, although both neutralising and non-neutralising antibodies contribute to defending the body.

The researchers found that five of the 13 antibodies neutralised the original strain. When they tested the neutralising antibodies against the new variants, all five antibodies neutralised Delta, three neutralised Alpha and Delta, and only one neutralised all four variants.

“In face of vaccination, Delta is relatively a wimpy virus,” Ellebedy said. “If we had a variant that was more resistant like the beta but spread as easily as Delta, we’d be in more trouble.”

The antibody that neutralised all four variants of concern—as well as three additional variants tested separately—was called 2C08. In animal experiments, 2C08 also protected hamsters from disease caused by every variant tested: the original variant, Delta, and a mimic of Beta.

Some people may have antibodies just as powerful as 2C08 protecting them against SARS-CoV-2 and its many variants, Ellebedy said. Using publicly available databases, the researchers discovered that about 20 per cent of people infected or vaccinated against SARS-CoV-2 create antibodies that recognize the same spot on the virus that is targeted by 2C08. Moreover, very few virus variants (.008 per cent) carry mutations that allow them to escape antibodies targeting that spot.

“This antibody is not unique to the person we got it from,” Ellebedy said.He added, “Multiple antibodies targeting this area have been described in the literature; at least one is under development as a Covid-19 therapy. Similar antibodies have been generated by people infected in Italy and people infected in China and people vaccinated in New York. So it’s not limited to people of certain backgrounds or ethnicities; it’s not generated only by vaccination or by infection. A lot of people make this antibody, which is great because it is very potent and neutralizes every variant we tested.”

Continue Reading

STUDY SUPPORTS WIDESPREAD USE OF BETTER MASKS TO CURB CORONAVIRUS INDOORS

Published

on

A new study highlights the need for widespread use of better face masks and the importance of good ventilation to mitigate the spread of COVID-19 indoors.

Engineering researchers at the University of Waterloo performed experiments using a mannequin to simulate a seated person breathing in a large room. The studies showed a significant buildup over time of aerosol droplets — exhaled droplets so tiny they remain suspended and travel through the air — despite the use of common cloth and blue surgical masks. “There is no question it is beneficial to wear any face covering, both for protection in close proximity and at a distance in a room,” said Serhiy Yarusevych, a professor of mechanical and mechatronics engineering and the leader of the study. “However, there is a very serious difference in the effectiveness of different masks when it comes to controlling aerosols.”

Previous research has found that aerosols dispersed by infected people are a source of transmission of the SARS-CoV-2 virus that causes COVID-19, even outside the two-meter social distancing zone widely recommended by public health officials.

The study showed that most common masks, primarily due to problems with fit, filter about 10 per cent of exhaled aerosol droplets. The remaining aerosols are redirected, mostly out the top of the mask where it fits over the nose, and escape into the ambient air unfiltered.

By contrast, higher-quality, more expensive N95 and KN95 masks filtered more than 50 per cent of the exhaled aerosols that can accumulate indoors and spread the COVID-19 virus when inhaled by other people.

Yarusevych, the principal investigator in the Fluid Mechanics Research Lab, said the much greater effectiveness of N95 and KN95 masks versus cloth and surgical masks makes a compelling case they should be worn in indoor settings, such as schools and workplaces, as much as possible.“A lot of this may seem like common sense,” he said. “There is a reason, for instance, that medical practitioners wear N95 masks — they work much better. The novelty here is that we have provided solid numbers and rigorous analysis to support that assumption.”

Experiments also quantified the impact of ventilation systems, which circulate and replace air in rooms, on the accumulation of aerosols. Even modest ventilation rates were found to be as effective as the best masks in reducing the risk of transmission.

Continue Reading

ANTIBODY PROTECTS AGAINST BROAD RANGE OF COVID-19 VIRUS VARIANTS

Published

on

The virus that causes COVID-19 today is not the same as the one that made people fall sick way back in December 2019. Many of the variants circulating now are partially resistant to some of the antibody-based therapeutics that were developed based on the original virus. As the pandemic continues, more variants inevitably will arise, and the problem of resistance will only grow.Researchers at Washington University School of Medicine in St. Louis have identified an antibody that is highly protective at low doses against a wide range of viral variants. Moreover, the antibody attaches to a part of the virus that differs little across the variants, meaning that it is unlikely for resistance to arise at this spot. The findings, published in the journal Immunity, could be a step toward developing new antibody-based therapies that are less likely to lose their potency as the virus mutates.

“Current antibodies may work against some but not all variants,” said senior author Michael S. Diamond, MD, PhD, the Herbert S. Gasser Professor of Medicine. “The virus will likely continue to evolve over time and space. Having broadly neutralizing, effective antibodies that work individually and can be paired to make new combinations will likely prevent resistance.”

SARS-CoV-2, the virus that causes COVID-19, uses a protein called spike to attach to and invade cells in the body’s respiratory tract. Antibodies that prevent spike from attaching to cells neutralize the virus and prevent disease. Many variants have acquired mutations in their spike genes that allow them to evade some antibodies generated against the original strain, undermining the effectiveness of antibody-based therapeutics.

To find neutralizing antibodies that work against a wide range of variants, the researchers began by immunizing mice with a key part of the spike protein known as the receptor-binding domain. Then, they extracted antibody-producing cells and obtained 43 antibodies from them that recognize the receptor-binding domain. Along with Diamond, the research team included co-first authors Laura VanBlargan, PhD, a staff scientist; Lucas J. Adams, an MD/PhD student; and Zhuoming Liu, PhD, a staff scientist; as well as co-author Daved Fremont, PhD, a professor of pathology & immunology, of biochemistry & molecular biophysics and of molecular microbiology.

The researchers screened the 43 antibodies by measuring how well they prevented the original variant of SARS-CoV-2 from infecting cells in a dish. Nine of the most potent neutralizing antibodies were then tested in mice to see whether they could protect animals infected with the original SARS-CoV-2 from disease. Multiple antibodies passed both tests, with varying degrees of potency.

The researchers selected the two antibodies that were most effective at protecting mice from disease and tested them against a panel of viral variants. The panel comprised viruses with spike proteins representing all four variants of concern (alpha, beta, gamma and delta), two variants of interest (kappa and iota), and several unnamed variants that are being monitored as potential threats.

One antibody, SARS2-38, easily neutralized all the variants. Moreover, a humanized version of SARS2-38 protected mice against disease caused by two variants: kappa and a virus containing the spike protein from the beta variant. The beta variant is notoriously resistant to antibodies, so its inability to resist SARS2-38 is particularly remarkable, the researchers noted.

Further experiments pinpointed the precise spot on the spike protein recognized by the antibody, and identified two mutations at that spot that could, in principle, prevent the antibody from working.

These mutations are vanishingly rare in the real world, however. The researchers searched a database of nearly 800,000 SARS-CoV-2 sequences and found escape mutations in only 0.04 per cent of them.

“This antibody is both highly neutralizing (meaning it works very well at low concentrations) and broadly neutralizing (meaning it works against all variants),” said Diamond, who is also a professor of molecular microbiology and of pathology and immunology.

Continue Reading