Why Hospital Door Handles Should Be Made Of Copper

This morning, when walking in from my car to my office in the hospital, I had to open 6 doors and press 2 elevator buttons. With each door handle and elevator button, I was having contact with someone else’s fomites. Fomites are the small droplets that are formed when a person coughs or sneezes and are the primary way that respiratory viruses get transmitted. Bacteria, such as Staph aureus, are instead usually transmitted by skin contact with infected tissues, for example, by touching body secretions containing bacteria. In both situations, it is the hands that are the primary conveyance mechanism – whether they touch a person infected with Staph or whether a person with the flu blows their nose into a hand-held Kleenex. Anything that those hands touch can spread the viruses or bacteria in those fomites. Those door handles and elevator buttons that I touched this morning contain the fomites of hundreds of patients and hospital staff.

Certain materials are inherently more antimicrobial than others. Most hospital door handles are made of stainless steel or aluminum because steel and aluminum wipes off easily and because it’s shiny appearance looks clean. However, it turns out that steel and aluminum do not have antimicrobial properties so viruses and bacteria can survive on steel knobs, buttons, and handles for hours. Copper, on the other hand, has potent natural antimicrobial properties. This has been known for centuries – for example, copper water vessels were preferentially used in ancient times because water would be of better quality when transported in copper containers as opposed to containers made of other materials.

The medical literature about copper’s antimicrobial properties is extensive and it has been shown to inhibit/inactivate fungi (such as Aspergillus and Candida), viruses (such as influenza and polio), and bacteria (such as E coli, Clostridium difficile, and MRSA).

One of the problems with copper as a building material is that it tarnishes and then looks old and dirty. However, copper can be combined with other metals to form alloys such as brass (copper + zinc) and bronze (copper + tin) which are more commonly used for building fixtures. Although these alloys are better than steel with respect to their antimicrobial properties, then are not quite as effective as pure copper. For example, 99.9% of E. coli are killed within 120 minutes on copper surfaces but similar kill rates take up to 270 minutes on brass surfaces and 270 minutes on bronze surfaces. In contrast, E. coli can survive on stainless steel surfaces for weeks. The antimicrobial properties of copper are temperature dependent. At temperatures of 4 degrees C (40 degrees F), bacterial killing requires 1 hour longer than at 20 degrees C (70 degrees F).

As we renovate our hospitals in the future, we should look back to using copper and brass door handles, cabinetry fixtures, and buttons. Copper may not look as clean and sterile as stainless steel but looks can be deceiving. A $4 door handle may just have as big of an epidemiologic impact as a $20,000 ultraviolet room sterilizer…

March 10, 2020



Why COVID-19 Mortality Is Higher In China Than The Rest Of The World

Initial reports from China in January 2020 was that the mortality rate of the novel coronavirus, COVID-19, was about 3%. Since then, the World Health Organization has been reporting daily mortality updates on its situation report website and the mortality rate has held at 3.4% in China. As of February 24, 2020, there have been 77,262 confirmed cases of COVID-19 in China with 2,595 deaths. However, outside of China, there have been 2,069 confirmed cases and only 23 deaths or 1.1%. So, why is the mortality rate of COVID-19 three times higher in China than anywhere else? When analyzing the outbreak, I believe that there is more than one reason.

Differences in hospital care

The United States gets a lot of criticism about poor overall national health compared to other economically advantaged countries. However, our nation’s poor health really relates to lifestyle issues (obesity, high numbers of suicide by gun, etc.) and outpatient healthcare (high percentage of uninsured citizens, poor access to primary care, high cost of medications, etc.). American inpatient hospital care is actually among the absolute best in the world. Similarly, Canadien and European nation hospitals provide great inpatient care to the sickest patients. On the other hand, many countries, such as China, lack the same ability to care for critically ill patients in their acute care hospitals and particularly in their intensive care units. The greatest danger of COVID-19 is respiratory failure and to successfully manage a large number of patients with respiratory failure, there has to be a well-trained population of critical care physicians, critical care nurses, and respiratory therapists. In addition, there has to be a sufficient supply of mechanical ventilators, monitoring equipment, and inpatient infection control practices in place. With SARS and Ebola outbreaks, the highest mortality rates were in those countries with the least ability to manage critically ill patients; in more economically advantaged countries, the mortality rates were much lower.

Virus detection rates

To accurately determine the mortality rate of a disease, you have to be able to identify every person who gets that disease. It is becoming clear that many patients with COVID-19 have milder forms of the disease and likely have gone undetected in China. In other countries, there have been efforts to track down and test any contacts of persons with confirmed cases of COVID-19 and as a consequence, milder cases are more likely to be detected in these other countries. When more of these milder cases are diagnosed, the ratio of deaths to confirmed cases falls resulting in a lower mortality rate. It is likely that there have been many more than 77,262 cases of COVID-19 in China and if that true number was known, then the mortality rate would likely be much less than the currently reported 3.4%.

Differences in the number of elderly and debilitated patients

In China, the outbreak of COVID-19 has struck large metropolitan areas, such as Wuhan (a city of 11 million people). Like influenza, the people who are most likely to die from COVID-19 infection are the elderly, the very young, and those people with chronic medical conditions. In cities such as Wuhan, there are large numbers of these vulnerable people living in close proximity to each other with the result that many elderly and debilitated persons become infected. Until this week, most of the cases outside of China have involved persons traveling from China. International travelers tend to be young and middle aged adults. Newborns, the very elderly, and those suffering from chronic disease tend to not do as much international travel. Those infected international travelers are therefore a select population of people who are best able to survive COVID-19 infection. This may be changing soon, with large numbers of domestically-acquired cases of COVID-19 being recently reported in South Korea and Japan.

Differences in quarantine practices

Here in the United States, the Centers for Disease Control has recommended a very high level of quarantine for patients with suspected COVID-19 infection including droplet, airborne, and contact isolation. This means gowns, gloves, negative airflow hospital rooms, and face shields. Many countries do not have the resources to provide this level of patient isolation. Indeed, persons suspected of having the infection are told to just stay home. In crowded, often multi-generational housing (such as exists in many Chinese communities), this quarantine practice can facilitate transmission to the vulnerable elderly, debilitated, and infants. In hospitals, sub-optimal isolation practices can facilitate transmission to other patients in that hospital and these patients with hospital-acquired COVID-19 are a select population of particularly vulnerable people who are more likely to die of the infection.

Lag-time differences

Death from COVID-19 does not occur instantly. There is first an incubation period of up to 14 days. Once a person develops symptoms, it takes several days to develop respiratory failure and die. The outbreak has been present in China for 2 months and so there has been sufficient time for those persons who get infected to develop symptoms and die of the infection. However, in other countries, patients with the infection have been identified and confirmed relatively early in their disease and in many cases, those people have not yet had time to develop respiratory failure and die.

Last year in the United States, an estimated 35 million Americans contracted influenza and 34,000 Americans died from influenza. That works out to about 1 death per every 1,000 people infected with influenza. The COVID-19 virus kills 34 out of every 1,000 people in China and about 11 out of every 1,000 people in countries other than China. If the outbreak spreads to the United States, based on these numbers we can expect COVID-19 to be about 10 to 11 times more lethal than influenza.

February 25, 2020


The Novel Coronavirus 2019-nCoV

In December 2019, an outbreak of a new coronavirus began in Wuhan City, China. Within weeks, it had spread throughout China and to a number of other countries. This post will outline the history of the outbreak, what we have learned from other, previous coronavirus outbreaks, and what physicians and hospitals need to know about managing suspected patients in order to control the outbreak.

Coronaviruses are common upper respiratory viruses that generally cause fairly mild infections. They account for about 5-10% of common colds and cause typical cold-like symptoms such as fever, cough, and sore throat. Like other upper respiratory viruses, they are spread by aerosolized droplets and primarily occur in the winter months. The regular coronaviruses are detected with the standard respiratory viral PCR panels that most hospitals use. These panels detect other common viruses such as influenza, rhinovirus, parainfluenza virus, and others. Treatment is supportive and there are no effective vaccines or anti-viral antibiotics. As with all respiratory viral infections, patients admitted to the hospital should be placed on droplet isolation.

Sometimes, Coronaviruses Go Rogue


From 2002 – 2004, an outbreak of a coronavirus resulted in SARS (Severe Acute Respiratory Syndrome). Overall, 8,098 people were known to become infected with 774 deaths, a 10% mortality rate. The natural host of the virus was the horseshoe bat and the outbreak was traced to a remote cave in Yunnan Province in China. From the bats in this cave, the SARS virus was spread to a masked palm civet, a cat-like wild animal in China that is often hunted for food. The civet was killed by a hunter and landed in a meat market in Guangdong, China in November 2002 where the virus then jumped to humans. Chinese health authorities were not forthcoming about the spreading Guangdong outbreak and did not report it to the World Health Organization (WHO) for several months, resulting in rapid spread of the infection due to a lack of public and healthcare worker awareness that it even existed. In February 2003, a businessman traveling from China became ill and was admitted to a hospital in Hanoi, Vietnam. An astute Italian physician working at the hospital, Dr. Carlo Urbani, recognized that the man’s infection was something different than regular influenza and notified the WHO. However, several healthcare workers at the hospital became infected; both the businessman and Dr. Urbani ultimately died of SARS. Meanwhile, a doctor from Guangdong unknowingly infected with SARS had traveled to Hong Kong where he stayed at the Metropole Hotel. He transmitted the virus to 16 international guests at the hotel who then traveled to Canada, Singapore, Viet Nam, and Taiwan, carrying the virus with them. On February 23, 2003, an elderly woman returning to Toronto from Hong Kong became ill with SARS and went to her local hospital. Canadien health authorities and healthcare workers did not have adequate infection control protocols in place and were unprepared for SARS and consequently within weeks, 257 people in Toronto were infected. In the United States, there were 27 cases and no deaths.

The incubation period for SARS ranged from 1-14 days but was usually 4-6 days. Like other coronaviruses, it was spread by droplets. With international efforts to stop the spread of the virus, the outbreak was contained and the last known case was in in January 2004. There are several lessons to be learned from SARS:


  1. Air travel permits rapid world-wide spread
  2. Don’t cover it up. By not being forthcoming about the emerging outbreak, Chinese authorities permitted the virus to quickly spread
  3. Rapid epidemiologic investigation is essential
  4. Countries must work together for epidemiological control
  5. Have a high clinical suspicion.
  6. Infection control works. Toronto did not have proper infection control practices in place, resulting in the Toronto outbreak
  7. Health personnel are at greatest risk


From 2012 – 2015, another coronavirus caused the Middle East Respiratory Syndrome. Overall, 1,227 people became infected with an average mortality rate of 37%. The natural host was the dromedary camel and the virus made the jump to humans in Saudi Arabia. There were two main outbreaks of MERS, the first in Saudi Arabia in 2014 when 402 people became infected. The second outbreak occurred in 2015 when a businessman traveled from the Arabian Peninsula to South Korea, unknowingly infected with the MERS virus. Ultimately, 150 Koreans became infected and the main locations that these people acquired their infections were hospitals and clinics that were not prepared to institute proper isolation precautions and infection control practices. Ultimately, the South Korean outbreak alone cost $8.5 billion. The main symptoms of MERS were cough, shortness of breath, and fever. Chest x-rays showed patchy pneumonia or ARDS patterns. There are several sessions to be learned from MERS:

  1. A single missed infected person can cause a nationwide outbreak
  2. Hospitals and ERs can accelerate spread
  3. Doctors in community hospitals and clinics are the first line of defense
  4. New coronavirus strains can have a very high mortality rate
  5. Outbreaks are expensive


Wuhan City is the 7th largest city in China with a population of 11 million. To put that in context, New York City has a population of 8.6 million, Los Angeles 4 million, and Chicago 2.7 million. Wuhan is a traditional manufacturing hub and the political, economic, and commercial center of Central China. Its location on the Yangtze River as well as its location at the intersection of several rail lines and highways makes it one of China’s main transportation hubs. The main rail station sees up to 80,000 people per day and the airport serves 20 million passengers per year with direct flights to Tokyo, Paris, London, San Francisco, and Chicago among other international destinations. In December, 2019, a new coronavirus called the “novel coronavirus” or 2019-nCoV appeared in Wuhan City and is believed to have originated at a Wuhan market where wild and domesticated animal meats are sold.

As with the SARS outbreak, Chinese health authorities were slow to recognize and report the emerging infection. The result was that the virus rapidly spread through Wuhan City and from there, to other towns and cities in China. Within weeks, the virus had spread to many other countries throughout the world, including the United States. Like other coronaviruses, the presenting symptoms are fever, cough, and shortness of breath. The incubation period is 1-14 days. In some people, the infection is rather mild but in others, it can result in severe illness. The mortality rate is about 3%.

When should 2019-nCoV be suspected?

  1. Fever AND symptoms of lower respiratory infection AND either:
    • Travel from Wuhan City, China in the past 14 days
    • Close contact in the past 14 days with a person under investigation for 2019-nCoV
  2. Fever OR symptoms of lower respiratory infection AND:
    • Close contact in the past 14 days with a person with laboratory-confirmed 2019-nC0V

ALL of the following isolation procedures should be used in suspected cases in the emergency department and the hospital:

  1. Contact Isolation:
    • Gown
    • Gloves
    • Hand hygiene with soap and water (alcohol based hand sanitizer if soap & water not available)
  2. Droplet Isolation:
    • Mask
    • Face shield or goggles
  3. Airborne Isolation
    • Negative airflow room
    • N-95 mask or PAPR

What should you do if you have a  suspected case?

  1. Place patient in isolation immediately
    • In the ER or hospital: contact + droplet + airborne islation
    • In an office setting: put patients with a possible history in a private room with a closed door; give the patient, family members, and healthcare workers a regular mask until additional history is obtained to determine if the patient requires transfer to a hospital for full isolation and additional testing
  2. Obtain a full travel history
  3. Communicate with the hospital infection control personnel and the local health department to determine if the patient needs testing
  4. Currently, testing is only done at the CDC and requires prior permission from the CDC

When performing testing, ALL of the following should be submitted:

  1. Sputum or bronchoalveolar lavage or tracheal aspirate
  2. Serum
  3. Nasopharyngeal AND oropharyngeal swab/wash/aspirate

Importantly, the standard respiratory viral panel test used by most hospitals does NOT detect 2019-nCoV. Therefore, a positive coronavirus test on a respiratory viral PCR panel does NOT indicate 2019-nCoV and instead indicates one of the regular coronaviruses that typically cause a common cold.

All testing should be performed with the patient in proper isolation, including airborne precautions; testing should not be performed in regular emergency department rooms or clinic rooms that are not capable of negative airflow. There is no effective anti-viral for 2019-nCoV so treatment is primarily supportive. For those patients who develop respiratory failure and require mechanical ventilation, intubation should be performed in a negative airflow room with all isolation precautions and should be performed by the most experienced physician available (this is not a procedure for trainees). If a person believes that he/she might have 2019-nCoV, then they should call ahead to the emergency department so that their throughput can be expedited and avoid exposing other patients in the waiting area.

The mainstays of response to any epidemiologic threat are preparedness, surveillance, containment, and education. Even though 2019-nCoV is frightening, influenza remains a greater threat to Americans. The CDC estimates that last year, 35 million Americans became infected with influenza and 34,000 died. So, even if 2019-nCoV is not present in your community, always use standard infection precautions:

  • Maintain proper hand hygiene practices
  • Cover your cough
  • Stay home if you are sick
  • Get your Influenza vaccination
  • Use droplet isolation with any admitted patient with a suspected viral respiratory infection
The 2019-nCoV outbreak is a rapidly evolving situation so check the CDC website and the WHO website for updates.
February 1, 2020

Predicting The Next Influenza Season

Last month, in August 2019, we have our first influenza case in our hospital. The patient recently returned from an international gathering so it is likely that he acquired the flu from a Southern Hemisphere resident attending the same gathering. The Southern Hemisphere has its influenza season at the opposite time of year as the Northern Hemisphere. A 2019 article in the journal Scientific Reports indicates that the epidemiology of influenza in Australia in any given season predicts the subsequent epidemiology of influenza in the Northern Hemisphere. So, what does the most recent Australian influenza epidemiology indicate?

The Australian Department of Health maintains a robust website with up to date influenza epidemiology information. The 2019 Australian data indicates that influenza started early this season as is depicted in the red line in the graph below.

Normally, influenza starts showing up in late May in Australia and peaks about the first of September. Australia started the onset of the 2019 influenza season in February, about 3 months earlier than usual, and it peaked about the first of July, about 2 months earlier than usual. So, let’s see how the Australian influenza data compares to information from the U.S. Centers for Disease Control influenza website:

In the above graph, we see the last 5 influenza seasons plus the 2009-10 season (the H1N1 epidemic) and the 2011-12 season. In 2009, the U.S. was caught off guard with the early arrival of H1N1 influenza that started in April and peaked in October. That year, 61,000 Americans got influenza and 12,500 Americans died of it (especially young persons). The 2017-18 influenza season (turquoise line in the graph) had the highest percentage of visits for influenza-like illness in the past 5 years. Comparing that curve to the 2017 Australian influenza season (brown line in the previous graph), there is remarkable similarity.

If the upcoming influenza season in the United States resembles the current influenza season in Australia, then it will start earlier than usual, perhaps in October rather than the normal December onset. The good news is that the most recent influenza season in Australia was not terribly severe with fewer hospitalizations and deaths than normal so hopefully this portends good news for U.S. hospitals and ICUs that may see fewer admissions and deaths from influenza.

What about the specific strains of influenza? Last year in the United States, influenza A H1N1 dominated early in the season in December (brown bars in the graph below). Influenza A H3N2 became the dominant strain by early March (red bars in the graph below). Influenza B was uncommon throughout the 2018-19 season until February and then it increased in prevalence when influenza A was decreasing (green bars in the graph below).



The graph below shows the various strains seen in Australia in the summer of 2019. Once again, influenza A peaked early in the season (red bars in the graph below) whereas influenza B peaked 4 months later (green bars in the graph below). Overall, influenza B accounted for a higher than usual percentage of the overall influenza cases in Australia. Influenza H1N1 all but disappeared shortly after the Australian flu season started and H3N2 was the dominant strain (although most of the Australian influenza A cases went unsubtyped). If the Australian experience predicts the upcoming flu season in the United States, then we should expect to see mostly influenza A H3N2 and also see a higher than normal number of influenza B this winter.

The 2019-20 U.S. standard strength influenza vaccine is quadrivalent and will cover the following 4 strains:

  1. Influenza A/Brisbane/02/2018 (H1N1)pdm09-like virus
  2. Influenza A/Kansas/14/2017 (H3N2)-like virus
  3. Influenza B/Victoria virus
  4. Influenza B/Yamagata virus

The vaccine is available in a standard strength that is indicate for persons over age 6 months and is made by several manufacturer, sold under brand names including Alfuria, Fluzone, Fluarix, and FluLaval. A high strength (Fluzone High Dose) contains about 4 times the amount of antigen and is indicated for persons over age 65. Unlike the quadrivalent standard strength influenza vaccine, the Fluzone High Dose vaccine is trivalent and does not cover the influenza B/Yamagata virus. This could have immunity implications if the U.S. sees an unusually high number of influenza B cases, as was seen in Australia this summer. Also, the Fluzone High Dose vaccine has a higher incidence of side effects such as injection site pain, fever, and myalgia. So the decision of which vaccine to use in persons age > 65 is not clearcut – the CDC does not recommend one vaccine over the other in people over age 65.

The good news for the upcoming influenza season is that we now have a new drug to treat influenza, baloxavir (trade name Xofluza®) that is given as a single one-time dose and will cost about $150. Tamiflu® is now available as generic oseltamivir at a cost of about $50 for a 5-day course.

So, what can physicians in the United States take away from this season’s Australian influenza season? Vaccinate your patients and do it now!

September 28, 2019