Pneumonia treatment considerations have changed over the years

June 2007

The etiologic agents of pneumonia in pediatric patients are age dependent but the likelihood of particular agents being involved has changed somewhat in the past few years.

In the normal newborn with early onset, group B streptococcus (GBS) was the major player. In the toddler, it was largely viral with the encapsulated organisms — Haemophilus influenzae type b (Hib) and pneumococcus — to “consider” and staphylococcus and group A streptococcus as outliers. Mycoplasma started to appear in school-aged children and gradually became a major consideration in the second decade. Chlamydia pneumonia occurred in the older end of the pediatric age group. The bacteria noted for toddlers were scattered through school-aged children and beyond.

There have been changes in the causative agents in recent years. These changes resulted from vaccines reducing the likelihood of Hib and pneumococcus, and prophylaxis against GBS decreasing early onset disease. The bad news is the emergence of resistant pneumococci and more recently staphylococcus.

Increasing methicillin-resistant Staphylococcus aureus has made the use of third-generation cephalosporins for initial treatment of pneumonia a bit chancy.

In the newborn period, early onset GBS has markedly decreased but is not gone.

When dealing with newborns, it is always appropriate to ask about perinatal history, particularly about premature rupture of membranes, maternal prophylaxis and also about previously affected infants. MRSA have shown up in nurseries, and many of these infections will present after the babies are discharged. It is a good idea to ask about what has been going around in the nursery. It always is appropriate to ask about a family history of skin lesions in a child with lower respiratory illness at any age. GBS late-onset pneumonia is less common than early onset.

Neonatal herpes simplex viruses (HSV) have received a lot of press but this disease still makes up only a couple of thousand cases annually in the United States. Pneumonia with no other involvement should not lead us toward a diagnosis of neonatal HSV. It is usually a component of systemic involvement, which will be accompanied by other findings, eg, elevated hepatic enzymes, encephalitis et al. It is well to remember, however, that systemic manifestations of neonatal HSV may precede the onset of skin lesions in some cases. A negative parental history of genital herpes is the rule rather than the exception.

Gram negative or fungal infections are usually a concern in newborns who have been hospitalized for some time. Finally, we should not forget that there are causes other than infection for tachypnea and dyspnea in full-term infants including anatomic and metabolic abnormalities. In any case, newborns suspected of having pneumonia, should have a blood culture, and an x-ray. I will typically pass on a lumbar puncture. Ampicillin and gentamicin are a good starting combination with appropriate therapy for staphylococcus when that diagnosis seems likely.

Most pneumonia in toddlers is viral. The WHO recommends tachypnea and retraction as simple signs for diagnosis of pneumonia.

Pneumococcal vaccine has achieved some reduction in pneumonia, and some have shown that this is more than could be explained on the basis of what we classically considered to be pneumococcal disease. While this organism accounts for a relatively small number of cases of pneumonia in this age group, it is a very significant group, as they require antimicrobial therapy whereas the majority of cases are viral and do not. There have been reports that some of the serotypes have been replaced by others not present in our current vaccine and this will bear watching. Hib vaccine, has practically eliminated invasive disease caused by this organism.

Bacterial, viral illness

How do we distinguish between bacterial and viral pneumonia? With difficulty. White counts above 20,000 should make us lean toward bacterial disease, but counts between 10,000 and 20,000 are not as compelling. There was a paper many years ago on white counts in proven viral croup, and counts of 15,000 were not unusual. This is not surprising to us now as we learn more about mediators of inflammation in infection. The other markers of inflammation appear to be just as good or bad as WBC for differentiation. Lobar consolidation is good evidence of bacterial infection. A positive blood culture is not the rule in bacterial pneumonia but is very helpful when positive.

Although we have encountered increased resistance to pneumococcus, most will respond to the levels achieved with a parenterally administered third-generation cephalosporins. However, for severely ill patients requiring hospitalization, it is prudent to initiate treatment with a third-generation cephalosporin and vancomycin. Vancomycin can be discontinued if sensitivity testing indicates it is not required and if the patient is recovering. Our reliance on third generations for initial treatment was shaken by the report that four children treated with what appeared to be appropriate therapy for pneumonia died from overwhelming community-acquired MRSA (MMWR. 1999;48:707). Reports of community-acquired MRSA have become much more widespread since this report.

A recent report from CDC of fatal staphylococcal pneumonia in children with influenza (Health Alert Network. Accessed online: May 9, 2007) is disconcerting. The average age of these children was 11 years. Thus MRSA must be considered as a complication during the influenza season when the clinical course is unusual. Antibiotics like nafcillin or oxacillin, although the choice for sensitive staph infections, will be of little use to us when dealing with MRSA. The initial choice would be clindamycin, which also is effective against pneumococcus, and group A streptococcus. A D test should be performed. Trimethoprim-sulfa is another option, although many strains of staphylococcus will exhibit multiple drug resistance. Vancomycin becomes the choice for initial in-patient therapy. Some MRSA are resistant to this drug as well. At this point, one should get consultation with an experienced pediatric infectious disease expert.

There has been much in the recent literature about the management of fluid collections in the thoracic cavity in pneumonia. These may represent effusions due to impairment of normal flow of pleural fluid from the visceral to parietal pleura in the course of pneumonia of whether one is dealing with empyema. Imaging will provide a lot of information on the location and amount of fluid as well as whether it is loculated and where best to drain it. In any case, there are good reasons for obtaining this fluid to differentiate these two conditions, to obtain an organism with which to better direct antimicrobial therapy and to remove mechanical inhibition of pulmonary ventilation. Decisions must be made as to whether one should simply remove fluid with a needle and syringe or drain the fluid though an indwelling tube. If there is empyema or continuing accumulation of fluid, tube insertion would be my choice.

Video assisted thoracic surgery (VATS) has been used very successfully in many centers to access loculated pus and take down adhesions. Children, in contrast to adults, have a remarkable ability to resolve intrathoracic damage from infection, and so pediatric surgeons tend to be more conservative in their approach to residual intrathoracic infection. However, VATS has been shown, at least in one center, to shorten hospitalization in children with pneumonia complicated by empyema.

Finally, we should not forget that mycoplasma pneumonia and Chlamydia pneumoniae are common causes of pneumonia after the first decade. Cold agglutinins is a relatively sensitive but nonspecific test for mycoplasma.

Specific serology for mycoplasma is preferred. While waiting, doxycycline or a macrolide should be started to treat both of these agents.

Pneumonia due to these organisms in normal children tends to be insidious, and resolution is also slow even when appropriately treated.

It is well to remember that etiology is usually age-related but seasonal and epidemiologic factors also need to be taken into account when managing children with pneumonia. Changes in antibiotic resistance will affect our choices, and availability of newer technology will assist in management.