Lori Kestenbaum, MD

is a fellow in Pediatric Infectious Diseases at The Children’s Hospital of Philadelphia. She graduated with a BS in Psychology from Duke University and received her MD from the Perelman School of Medicine at the University of Pennsylvania. She completed her residency in Pediatrics at The Children’s Hospital of Philadelphia in 2012.  She is currently a member of the American Academy of Pediatrics, the Pediatric Infectious Diseases Society and the Infectious Diseases Society of America. Follow her on Twitter @lorikestenbaum.

First, Do No Harm

As a first year infectious diseases fellow, I am frequently challenged to give a child an infectious disease diagnosis prior to proposing starting antibiotics. It brings me back to medical school, when we were taught to name the diseases, then the bugs that can cause that disease, then the drugs that can treat those bugs. I went into ID because of how neatly that thought process can be carried out, even in the most complicated patients.

However, every once in a while, that thought process breaks down, and our patients suffer. I consulted on a child this year with toxic epidermal necrolysis (TEN), which is a severe, often fatal, adverse cutaneous drug reaction that predominantly involves the skin and mucous membranes. She had had a few illnesses in a 3-month period which prompted a flurry of doctors’ visits, including multiple visits to the pediatrician, one visit to seek out a second opinion and one visit to an infectious diseases consultant.

The pediatrician had felt these were a series of viral illnesses; the second physician ran some blood work, a urinalysis and a urine culture. The urinalysis had 6-10 white blood cells, and Ah ha! the diagnosis had been made: a urinary tract infection. The doctor started trimethoprim-sulfamethoxazole. Two days later, the culture returned with no growth. The family was called and told that no organism had been identified, but the child might as well finish the course of the antibiotic.

The following week, the child developed a small cough, followed by injected eyes. As is described in the worst cases, she had mucocutaneous tenderness, hemorrhagic erosions, erythema and more or less severe epidermal detachment which led to blisters and areas of denuded skin, which involved upwards of 90% of her body surface area. She had the most severe case of TEN my attending could remember due to unnecessary exposure to trimethoprim-sulfamethoxazole. She required prolonged intubation while her lungs and airway healed, ophthalmologic surgery, meticulous nursing care and many weeks of rehabilitation. Her mother relayed the phone conversation she had when the antibiotic was continued, even though her urine culture was negative. “I don’t know why we kept giving her the antibiotic when she didn’t have an infection.” I couldn’t agree more.

Antibiotics have a reputation as being mild, potentially harmless medications that can only bring benefit. When the right drug is used for the right bug, they can be nothing short of lifesaving. When needed, we accept the risks and adverse effects than come with any medication. However, when a drug is used and there is no bug, they cannot benefit the patient and can only bring harm.