Current Perspectives on Pneumonia in Clinical Practice

[Guest guest]

From Medscape Pulmonary Medicine

Expert Interview

Current Perspectives on Pneumonia in Clinical Practice

07/14/2008

Medscape*

S. Niederman, MD

http://www.medscape.com/viewarticle/577190

Author Information

Information from Industry

Assess clinically focused product information on Medscape.

Click Here for Product Infosites -- Information from Industry.

Introduction to Perspectives on Pneumonia

The increased concerns of drug resistance in managing lung disease

affects both hospitalized and outpatient populations. In this expert

interview with Niederman, MD, of State University of New

York and Winthrop University Hospital conducted by Medscape editor

Brown, PhD, the risks for pneumonia in special populations

are considered. The current management trends to prevent and treat

pneumonia caused by ever-evolving pathogens are put into a clinical

perspective in the following commentary.

Patients at Risk for Pneumonia

Medscape: Welcome, Dr. Niederman, and thank you for sharing your

perspectives on pneumonia. What patient populations are currently

considered to be at increased risk for bacterial, viral, or fungal

pneumonia?

Dr. Niederman: All 3 forms of pneumonia can affect virtually any

individual, but in general the paradigm for any type of pneumonia is

the balance between the patient's host defenses, the virulence of

the potential pathogen, and the size of the exposure to the

pathogen. So in theory, pneumonia can develop even in healthy

individuals if they encounter particularly virulent pathogens or

very large inoculum. So for example, an individual who gets severe

acute respiratory syndrome (SARS) encounters a virus so virulent

that it is impossible to have host defenses against it. On the other

hand you can have a large inoculum of bacteria, which could

overwhelm a healthy defense system. A good example is a young,

otherwise healthy individual who gets drunk, vomits, and aspirates

such a large quantity of bacteria that pneumonia develops. The

majority of the patients we see in the hospital who have pneumonia

tend to be individuals with impaired host defenses, either as a

result of advanced age, underlying comorbid medical conditions, or

medications. Because of their impaired immune systems not only are

these individuals more prone to infection, but when pneumonia

develops, it is a more severe infection.

Comorbid conditions could be congestive heart failure, chronic

obstructive pulmonary disease (COPD), cigarette smoking, or

underlying malignancy. A variety of medications that we routinely

use increase the risk for infection. Even something as simple as

aspirin has been shown to interfere with immune defenses and can

predispose some patient populations to pneumonia. Certainly advanced

age is also an important factor. There is some controversy about

whether it is the age itself or the diseases that develop as a

result of getting older. It is probably to some extent the

combination of both.

Medscape: Are there special concerns for the immunocompromised, for

example, those who take medications that are immunosuppressive?

Dr. Niederman: There are concerns, and the most common

immunosuppressive drugs of concern used in clinical practice are

probably corticosteroids. Individuals who are chronically treated

with corticosteroids, as might be the case in those with COPD,

rheumatoid arthritis, or other inflammatory diseases, are more prone

to bacterial and viral infections as a result of their steroid

therapy. Chemotherapy immune-related suppression can develop in

individuals receiving chemotherapy, which particularly predisposes

them to bacterial and fungal infections. With the biologic therapies

that interfere with cell-mediated immunity, particularly the anti-

tumor necrosis factor types of therapies, there has been a much

greater concern about tuberculosis than pneumonia. Anything that

interferes with immune defenses can predispose a patient to both

bacterial and fungal infections.

Medscape: In patients with, for example HIV or a disease that has

made them immunocompromised, what are the particular risks for

pneumonia?

Dr. Niederman: There is a risk for immunocompromised patients. The

way to evaluate that risk is to look at the abnormality the patient

has to know, for example, what that does to the immune system. Then

you can better predict what types of pathogens are likely to cause

infection. So for example, people who are infected with HIV can have

predominantly abnormalities of cell-mediated immunity but they can

also have antibody formation abnormalities. Therefore, they are on

the one hand predisposed to things like tuberculosis and atypical

microbacteria and unusual fungal infections. In addition, they're

also greatly at risk for pneumococcal pneumonia.

If you look for example at patients who have neutropenia because of

cancer and chemotherapy, they are particularly predisposed to Gram-

negative bacteria. So if you can define the nature of the

immunosuppression and where specifically it affects the immune

system, this can help you better predict which specific types of

pathogens are likely to be problematic.

Evaluating Causes of Pneumonia

Medscape: In patients with cystic fibrosis in particular, what are

the most common causes of pneumonia you see?

Dr. Niederman: The 2 pathogens that have been a problem in cystic

fibrosis are Staphylococcus aureus, which tends to be a pathogen

earlier in the disease and Pseudomonas aeruginosa, which tends to be

a pathogen later in the disease. But there are a variety of unusual

Gram-negative bacteria that have been a problem with cystic fibrosis

as the disease gets more advanced. Here the factors that predispose

patients to infection tend to be the cystic fibrosis itself, which

leads to alterations in production of mucous, quality of the mucous,

and clearance of the mucous. All of these can predispose to specific

bacteria. Inevitably because of their disease, patients with cystic

fibrosis are frequently treated with antibiotics. The treatment can

then predispose to emergence of particularly resistant forms of Gram

negative bacteria that tend to be problematic later in the disease.

Medscape: How would you determine the cause in a patient who has

symptoms of pneumonia, whether bacterial, viral, or fungal?

Dr. Niederman: I think one of the issues that people are struggling

with right now is this question, is it possible to look at a patient

and know whether their infection is bacterial or viral? Efforts in

the past included can you tell by the clinical syndrome, can you

tell by the radiographic patterns, by the fever, by the white cell

count? Many studies have been done but they have not yet shown that

clinical features can separate bacterial from viral infections. One

of the reasons this is important is if you can predict who has a

viral infection with some accuracy, then you could say that this

person doesn't need antibacterial therapy. Since we are concerned

about the overuse of antibiotics and how that could be driving

resistance, it would be helpful in theory to be able to predict who

has which.

A particularly promising approach right now involves serum

measurements of procalcitonin.[1] When procalcitonin measurements

are low, they exclude bacterial disease, as low procalcitonin levels

seem to occur in viral infection. Studies in patients who have x-ray-

positive pneumonia and low procalcitonin levels who have been

randomized to observation without antibiotics have done very well

without antibiotic therapy. This is one potential measurement for

the future.

Today, for most patients, we look at their clinical features. If

there is any chance that it is bacterial, we treat them with

antibiotics and we try to identify the bacterial pathogen. Sputum

samples are used if patients are intubated and the tracheal

aspirates or bronchoscopic samples are cultured. We examine them

microscopically to identify suspicious bacterial infections, and to

identify what bacteria.

Medscape: Are agent-specific tests recommended for regular use?

Dr. Niederman: Generally agent-specific tests are not routinely

used. Agent-specific tests widely available right now include

urinary antigen testing for Legionella and Pneumococcus species. The

problem with specific-agent tests in community-acquired pneumonia

(CAP) is that it's difficult to know when the test is positive and

how it should affect your approach to treatment. There are some new-

onset cases and complexities of treatment. For example, there is a

dataset of patients who have been in infected with Pneumococcus, who

have pneumococcal bacteremia, who seem to benefit by the addition of

a second antibiotic. If that is reproducible, then it would be hard

to imagine how doing a specific test for Pneumococcus that was

positive might change the antibiotic therapy. It might narrow it if

you have a broad-spectrum therapy because of the possibility of

needing a second antibiotic. But Legionella and Pneumococcus urinary

antigen tests are probably the most specific tests in use.

The other common approach is to culture sputum, tracheal aspirates,

bronchoscopic samples, or lower respiratory tract samples in

patients who have suspected or confirmed pneumonia.

Current Practice for Treating Pneumonia

Medscape: What are the current medications of choice and how is the

therapy selection individualized for the patient newly diagnosed

with pneumonia?

Dr. Niederman: Guidelines for treating patients with CAP, with

healthcare-associated pneumonia (HAP), and with nosocomial pneumonia

depend on the severity of illness, the comorbidities of the

patients, and the time of onset of the illness in the case of

nosocomial infection. There are a variety of different algorithms

for therapy, but no single best therapy. The 2005 guidelines for

nosocomial pneumonia were a joint effort of American Thoracic

Society and the Infectious Disease Society of America, as were the

2007 guidelines for CAP.[2,3]

In general, with CAP, we treat for pneumococcus and generally the

possibility of atypical pathogen co-infection. Some subpopulations

are treated for Gram-negative bacteria. For patients who have HAP,

coming from a nursing home or the hospital in the previous 3 months,

they are at much greater risk for drug-resistant Gram negative

infections and methicillin-resistant S aureus, and these patients

need therapies directed to those pathogens. In nosocomial pneumonia

it depends on how late in the course of the hospital stay pneumonia

developed. Have they recently been on antibiotics? What are the

pathogens that are common in a given hospital? All of these factors

then get considered in terms of choosing an antibiotic therapy.

Strategies for Preventing Pneumonia

Medscape: Considering care of the elderly, could you please comment

on strategies to prevent pneumonia in this population?

Dr. Niederman: I think in the elderly, 3 factors may play a role in

their increased risk for infection. One is aging itself, and there

is some controversy whether that really plays a role. The

medications that the elderly get, and also the comorbidities that

they have both play a role. For the elderly, many medications and

comorbidities are associated with the risk for pneumonia. In terms

of prevention, prevention of pneumonia generally involves smoking

cessation and vaccination for patients outside of the hospital

setting. For the elderly individual who is residing outside of the

hospital, whether it's in the nursing home or at home, vaccination,

including both pneumococcal and influenza vaccines, are probably the

2 most important preventive strategies that we have right now. In

the hospital a whole series of preventive measures is undertaken.

Medscape: A recent report indicated vaccinating even the young had a

positive effect for preventing pneumonia in others with HIV in the

population.

Dr. Niederman: This is an interesting observation, the concept of

herd immunity. When the 7-valent conjugant vaccine for Pneumococcus

[4] (PCV7) was widely used in children, it had a benefit in elderly

individuals, particularly the ones caring for the children. It

seemed that if you eliminated the reservoir of Pneumococcus in young

children, you might then eliminate the distribution source, if you

will, of the Pneumococcus to older individuals who came in contact

with the children. Very compelling data showed that with the early

use of the vaccine in children there was a decline in the frequency

of those strains of Pneumococcus in elderly individuals. This

paralleled the decline in children, and the speculation was that

they were therefore not transmitting the Pneumococcus at the same

rate to the elderly caregivers around them.

Another twist to that data on vaccinating the young should be

watched very closely. There have now been reports that with

continued use of PCV7, there appears to be a decline in the

frequency of the strains covered in that vaccine. They have been

replaced at a higher frequency by other strains that aren't covered.

In particular there are some very severe, necrotizing strains of

Pneumococcus that aren't covered by the current vaccine that have

emerged at higher frequency in vaccinated individuals. The concern

is that the Pneumococcus being able to adapt to this vaccination

strategy has now replaced some of the vaccine strains with other

strains, some of which are more virulent, and that effect may be

that there is more severe pneumonia occurring in certain populations

than was appearing before the vaccine was introduced.[5] So if that

turns out to be true and can be confirmed in other populations, it

may negate some of that herd immunity benefit that has been seen in

the nonvaccinated individuals.

Medscape: What preventive measures would you say are helpful in the

hospital setting to reduce HAP?

Dr. Niederman: Well the big issue right now in the hospital is the

focus on mechanically ventilated patients, and there a lot of

efforts have revolved around something that has been referred to as

a " ventilator bundle. " It's a bundling of interventions that many

hospitals do in an effort to reduce the frequency of pneumonia. The

typical bundle most widely promoted and used involves 5 elements,

two of which are part of the bundle but really are good patient care

that can help reduce intensive care unit stay and may indirectly

reduce pneumonia but don't directly interfere with pneumonia

pathogenesis. Those are gastrointestinal bleeding prophylaxis and

deep venous thrombosis prophylaxis.

The other 3 interventions in the bundle are much more pneumonia-

specific. One is elevation of the head of the bed with the idea of

minimizing gastric reflux into the oropharynx and then aspiration

into the lung. Keeping the head of the bed elevated can reduce

gastric transmission of bacteria to the lungs. The 2 others are a

daily interruption of sedation -- you wake the patient up every day,

and when you do that there is a daily effort of weaning. That is the

fifth element. So the bundle is: daily interruption of sedation and

daily weaning trials, head of the bed elevation, deep vein

thrombosis and gastrointestinal bleeding prophylaxis.

Although these bundles are probably effective, their advent has led

to some interesting discussions and controversies. Some individuals

are reporting that with the use of ventilator bundles they have

eliminated pneumonia from their hospital. Some hospitals have even

said we have gone as long as 2 years without an episode of

ventilator associated pneumonia because of how valuable these

bundles are and how successful we've been in implementing them.

The problem with those reports has been that many studies report a

lower frequency of pneumonia in ventilated patients, but don't seem

to report secondary benefits, meaning overall reduction in use of

antibiotics, reduction in mortality, reduction in length of stay --

the known consequences of VAP. The diagnosis of VAP can often be

subjective, and it's an elusive, difficult diagnosis. The worry is

that people are now saying they don't have pneumonia in their

hospital, but it isn't as credible as you would like because the

secondary consequences of pneumonia are not also disappearing.

All of this has become increasingly more controversial because

Medicare has proposed, and it's unclear what the final decision will

be, that VAP should be considered a potential medical error that

should not be given additional reimbursement. If so, the concern is

that many hospitals would report very low rates of pneumonia.

However, it's not clear to me that they're really eliminating the

problem. It seems like there are populations of patients who are at

such high risk because of their underlying comorbidities, chronic

illness, and acute illness that they can't possibly have a zero-risk

for pneumonia, even in the best hands and in the best hospitals.

Medscape: That really leads us directly into the Hospital Compare

Web site. What impact will public reporting have on practice, or how

will it change how patients choose hospitals?

Dr. Niederman: The Hospital Compare Web site right now is looking at

CAP, but it may become part of the Medicare database that they'll be

publicly reporting on VAP rates, and I would make the distinction. I

think that in general it is good that pneumonia has become a target

of public reporting, that people are aware of the importance of

pneumonia. It is the number one cause of death from infectious

diseases in the United States, and we are doing everything we can to

reduce the frequency of pneumonia.

But you have to wonder whether the public reporting can have some

unintended consequences that are not positive. One worry in the VAP

side is if we have public reporting of very low VAP rates, that sort

of strains credibility. There are populations where it's impossible

to get the rates as low as some people report; you wonder if they

are simply redefining the illness and haven't really made progress.

What I worry about personally is that if we say pneumonia is a

medical error, VAP, and it really is not eliminated, it will really,

I think, hamper future research in this area. It would be very

difficult for people to want to investigate a disease that when it

occurs, it is considered bad medical care. So I think we have to get

very accurate and honest in defining how low we can reduce these

rates.

The Hospital Compare Web site, which compares different hospitals

and their compliance with CAP, in general has been helpful. The one

area that I think it still controversial is the antibiotic timing

measure that Medicare has instituted. Originally that was looking at

the frequency with which antibiotics were given within 4 hours and

that has now been changed to 6 hours. But the concern has been that

if it becomes publicly reported that 4 hours is the standard, then

hospitals will inevitably try to get that rate to 100%. I think that

many experts believe that 100% compliance with the 4-hour rule isn't

necessarily good medical care. There may be cases of medical

uncertainty, and so as another modification they have introduced

medical uncertainty as an exemption to this 4-hour rule.

Published reports have raised concern about the 4-hour antibiotic

timing rule. There are now 2 retrospective studies that show that

when hospitals try to comply with the 4-hour rule, in general they

give antibiotics to more patients, but the frequency with which the

antibiotics were used for a final discharge of CAP actually dropped.

In other words, many more patients got antibiotics but some of them

got antibiotics when they didn't really have pneumonia.

Another consequence is at least one report of a hospital with very

bad episodes of infection with Clostridium difficile. The majority

of the episodes occurred in patients who were treated with

antibiotics for pneumonia, but half of those patients who were

treated for pneumonia didn't, in retrospect, have pneumonia. This

occurred at a time period when the hospital was trying to be more

compliant with the antibiotic timing measure.

I think that the Hospital Compare Web site is generally a good idea,

but the goals that are set in comparison have to be accurate and

realistic. What I worry about is that when the data get published

publicly, everyone is going to look for the hospital that has the

highest number. You have to recognize that it might be possible for

some of these measures, that the highest number may not necessarily

equal the desired goal. The desired goal might be 85% rather than

100% for some of these measures. It might be that there is a bell-

shaped curve where the best care occurs when the compliance rate for

a hospital is 80% to 85%. People below that and above that might not

be providing the best possible care.

Medscape: Thank you for sharing your time, Dr. Niederman.