News & Updates

Dr. Paul Kamitsuka Provides Lecture Notes on the Clinical Approach to Infections in the Elderly

March 04, 2010

Paul F. Kamitsuka, M.D., D.T.M.&H.

Harvard Medical School Postgraduate Course in Geriatric Medicine

February 2010, Boston, MA

Clinical Approach to Infections in the Elderly

Unique Aspects of Infection in the Elderly

· Infection in the elderly is associated with increased morbidity and mortality compared to younger age groups, and this is true for virtually all infection syndromes. For example, 90% of mortality due to influenza and pneumonia in the U.S. occurs among those 65 years and older.

· Typical signs and symptoms of infection are frequently absent in elderly debilitated patients. Infection should be suspected when there is (1) a decline in functional status such as new of increasing confusion, incontinence, falling, deteriorating mobility, anorexia, or failure to cooperate with nursing facility staff; (2) fever, defined as a single oral temperature of > 100° F or repeated oral temperatures of > 99° F or rectal temperatures of > 99.5°F, or an increase in temperature of > 2°F above baseline.¹ Delays in diagnosis and therapy due to subtle or even misleading clinical presentations may contribute to increased morbidity and mortality.

· The functional and residential status of the patient may determine the flora of concern. For example, pneumonia in the nursing home setting may be due to gram-negative rods and S. aureus in addition to the "usual" flora of community-acquired pneumonia. Moreover, resistant gram-negatives and MRSA are more common in the nursing home population and must be considered when choosing empiric antibiotics for these patients.

· Finally, advanced age, comorbidity, and malnutrition may reduce immune responses to vaccination.^2,3 Despite reduced efficacy, vaccination is still beneficial and grossly underutilized.

Influenza.

· 2009 was historic with respect to influenza with the emergence of the 2009 H1N1 pandemic, the first global influenza pandemic in 41 years. Since 2009 H1N1 descended from past pandemic viruses (1919, 1957, 1968) to which the elderly may have previously been exposed, many elderly persons harbor cross-protective immunity against this virus. Thus, the brunt of mortality due to 2009 H1N1 (89% by mid 11/09, CDC) was borne by younger age groups. This is in marked contrast to inter-pandemic years where 90% of influenza-related deaths occur among the elderly, usually by exacerbating underlying co-morbid conditions such as heart and lung disease, with a yearly average of 36,000 deaths and 225,000 hospitalizations in the United States.⁴

· The 2009 H1N1 pandemic highlighted critical vulnerabilities with respect our ability to diagnose, prevent and treat influenza. Rapid influenza tests were shown to perform poorly for this infection^4a-4c, the release of vaccine was painfully delayed due to antiquated production methods and poor vaccine yields^4d, available antivirals proved only modestly effective, and antiviral resistance threatened our therapeutic capability.^4e It is sobering to note that there are no FDA-approved antiviral drugs effective against influenza that may be given by either the intravenous or intramuscular route.⁵ Moreover, the same point mutation,

H275Y (histidine to tyrosine at amino acid 275 on the neuraminidase gene) that has rendered almost all seasonal H1N1 influenza strains resistant to oseltamivir (Tamiflu) throughout the world in just the past three years is also emerging among pandemic 2009 H1N1 isolates. Oseltamivir-resistant isolates are less susceptible to peramivir, a parenteral antiviral currently in development. Many elderly persons, particularly those with pulmonary disease or those unable to inhale on command, cannot use zanamivir (Relenza), the only remaining effective drug.

Unfortunately, there are no new anti-influenza agents on the near-term horizon⁵.

· Given this circumstance, it is critical that we re-think our approach to protecting the elderly from influenza, and maximize methods currently at our disposal.^5a Chief among these, particularly in the absence of truly effective antiviral therapy, is the optimal and widespread use of influenza vaccination. It is critically important to achieve > 90% vaccination rates among our patients, our healthcare staff, and among family members to whom the elderly are exposed. Current and evolving infection control guidelines for healthcare facilities, too numerous to enumerate here, address the elimination of potential exposures, engineering controls, administrative controls, and the use of personal protective equipment⁶. With the threat of oseltamivir resistance, the use of oseltamivir prophylaxis needs to be carefully reconsidered. Liberal use of oseltamivir prophylaxis in LTCF in response to influenza outbreaks, as has been practiced in the past, may save lives in the short-term but may also facilitate oseltamivir resistance that may further jeopardize the utility of this drug in the near future. Far more emphasis needs to be placed on making sure that infection control measures are practiced and enforced, and that high vaccination rates are achieved, creating a protective cocoon around patients to lessen the chance of outbreaks occurring in the first place.

· Vaccinating Our Patients. Annual influenza vaccination is the most effective measure to prevent influenza and its complications⁴. While the influenza vaccine may not be particularly effective in preventing influenza illness in the elderly, especially when there is a mismatch between the circulating influenza strains and the vaccine components⁷, it still provides significant protection against morbidity and mortality. A study published in 2007 underscored this point⁸. Data were pooled from 18 cohorts of community-dwelling members >65 years of HealthPartners in MN and WI for 1990-1991 through 1999-2000, and two other HMOs (Kaiser Permanente Northwest in Portland, OR and Oxford Health Plans in NYC and surrounding counties) for 1996-7 through 1999-2000. There were 713,872 person-seasons of observation. Vaccination was associated with a 27% reduction in the risk of hospitalization for pneumonia or influenza (adjusted odds ratio, 0.73; 95% CI, .68 to .77) and a 48% reduction in the risk of death (adjusted OR, .52; 95% CI, .50 to .55). Interestingly, a recent paper suggested that cardiovascular exercise training in healthy but sedentary older adults may result in a significant increase in seroprotection 24 weeks after vaccination.^8b Among those living in nursing homes, available data suggests that influenza vaccine may be as much as 80% effective in preventing death, although its efficacy in preventing influenza illness is only in the range of 20-40%⁴.

· How well are we doing in reaching influenza vaccination targets in the elderly? National Health Interview Study (NHIS) data from 2006 estimated influenza vaccine coverage among persons aged ³ 65 years to be only 66%, far short of the 90% national goal ⁹. This rate of vaccination was unchanged since 1997. Moreover, there is a significant racial/ethnic disparity in vaccination rates. Coverage among persons ³ 65 years in 2005 was 68% for non-Hispanic whites compared to only 47% for non-Hispanic blacks and 49% for Hispanics. Access to health care is an important but only partial explanation for the racial/ethnic disparity noted. ¹⁰ Factors such as patient attitudes and provider performance may also contribute to this disparity. Vaccination rates in nursing homes also fall short. In a recently published study from Michigan, only 57-61% of nursing homes vaccinated > 80% of residents with mean vaccination rates of 76-79%.¹¹ Vaccination rates of 80% or more in nursing homes have been shown to provide herd immunity protection and reduce the risk of outbreaks.¹²

· Vaccinating health care providers (HCP). Simply stated, if less than 90% of those working in your health care facility are not being vaccinated against influenza each year, not enough is being done to protect your patients. Vaccinating HCP against influenza is the single most effective measure to prevent transmission of influenza within health care facilities.^{13, 13a} Influenza cannot develop among hospitalized patients or long-term care residents if they are not exposed to influenza virus from others – namely HCP and visitors. Unfortunately, HCP vaccination rates remain appallingly low – only 42% ¹⁴ and similar to the rate for the general U.S. population where < 40% received the 2008-09 influenza vaccination.⁴ Vaccinating HCP not only benefits the vaccinee by preventing illness and absenteeism,^{15, 16} but because of herd immunity, high vaccination rates extend protection to highly vulnerable groups like the elderly who may have inadequate responses to vaccination.¹⁷ During average flu seasons about 15% of persons become ill, but illness rates in HCP may be as much as double this rate. In one nosocomial outbreak, for example, 35% of HCP became ill.¹⁸ A survey showed that > 75% of HCP with an influenza-like illness continued to work. Such persons spread influenza to co-workers and to vulnerable patients who might then suffer a fatal outcome as a result. Infected adults shed influenza for at least 1 day prior to the onset of symptoms, and for a total of 5-10 days. Vaccinating HCP against the flu reduces mortality among LTCF residents.^19-21 A recent study by the Rand Corporation underscores these points.²² The authors examined influenza vaccination rates in 301 nursing facilities in a single for-profit chain during the 2004-2005 flu season, comparing vaccination rates with rates of influenza-like illness (ILI) clusters reported from these facilities. The key finding was that both staff and residents must have high rates of vaccination to substantially alter the rate and impact of influenza outbreaks. Facilities having greater than 55% of staff and greater than 89% of residents immunized were almost 60% less likely to have an ILI cluster (OR: 0.410; 95% CI 0.19-0.89) compared to facilities with lesser rates.

· How do we increase HCP influenza vaccination rates? As noted above, achieving > 90% vaccination among staff members should be a top priority in all healthcare institutions. This goal can only be achieved with concerted and organized effort among institutional leaders to make influenza vaccination an expectation as a matter of safety and quality of care. Staff education, free vaccine for employees, and systems enabling easy access to vaccination (e.g., nurses vaccinating each other on the floor, roving vaccination carts, etc.) are critical components of successful programs. Misconceptions regarding influenza vaccination abound among HCP, as well as members of the general public²³ who choose not to be vaccinated.^23a Some HCP continue to believe that they can contract influenza from flu shots, even though this is biologically impossible. Others object to the presence of thimerisol, or fear that vaccines may be linked to autism, even though there is no credible evidence to support this concern.^{4, 23b} Misconceptions about vaccine safety are common, and must be corrected. Importantly, many HCP are simply unaware of the considerable health risk they pose to patients if they are not vaccinated. Some healthy persons may have asymptomatic, or minimally symptomatic, illness due to influenza but are still contagious to others, so that their minor illness when transmitted to an elderly person with multiple co-morbidities may actually lead to their patient's demise. It's clear that we must create an atmosphere of accountability among healthcare staff, taking concrete steps to ensure that staff vaccination rates are high.^4,13 Several organizations, including the Infectious Disease Society of America, recommend mandatory vaccination for HCP, with a declination provisions for religious or medical reasons.²⁴ Likewise, several states have moved toward mandating influenza vaccination for HCP.²⁵ The constitutionality of vaccination mandates has been questioned. Court decisions, including that of the U.S. Supreme Court in Jacobson v Commonwealth of Massachusetts, have historically upheld measures compelling individuals to accept vaccinations in order to protect the public health.^{26, 27} Short of individual mandates, however, there is much that can be done to urge, cajole, and compel HCP to be vaccinated, and exceeding 90% influenza vaccination rates should be a yearly institutional mandate.

· Recommend that all visiting family members, especially children, receive influenza vaccination. Increasing vaccination rates among visitors to LTCF may lower the risk of influenza for the debilitated elderly.²⁸ Unlike adults, young children may shed influenza virus for several days prior to illness and can be infectious for 10 days or more after the onset of illness. As a matter of patient safety, strong consideration should be given to restricting visitation to those able to document their having received the influenza vaccine. Clearly, those with signs or symptoms of clinical illness during influenza season (visitors and staff alike) should avoid coming to the facility. Other recommendations may be found at the CDC website (www.cdc.gov/flu) and elsewhere.^28b

· Infection Control Measures for Influenza. Direct or indirect contact with respiratory droplets is the primary means by which influenza is transmitted. Stringent hand hygiene may reduce the spread of this infection in healthcare settings. Handwashing with soap and water and the use of alcohol-based hand rubs both eliminate influenza virus from contaminated hands.²⁹ The issue of what kind of masks to use when caring for patient on isolation for H1N1 influenza has been controversial. The Institute of Medicine ^20b, the CDC, and OSHA (whose representatives enforce based on CDC guidelines) recommended the use of N95 respirators by those entering the rooms of patients with suspected or documented influenza infection. However, investigators in Ontario found standard surgical masks to be non-inferior in preventing acquisition of laboratory-confirmed influenza among nurses caring for influenza patients.^20c Similar findings were found by investigators in Australia.^20d Based on these and other data, the Infectious Disease Society of America (IDSA), Society for Healthcare Epidemiology of America (SHEA), and the Association for Professionals in Infection Control and Epidemiology (APIC) all recommended that droplet precautions (standard surgical masks, single rooms, handwashing) be practiced when caring for patients with suspected H1N1 infection, and that use of N95 mask be reserved for procedures such as bronchoscopy, open suctioning, and intubation. This controversy was further exacerbated by a nationwide shortage of N95 masks during the past several months.

Pneumonia in the Elderly

· Pneumonia is the leading infectious cause of death in North America.³⁰ As with other infections, pneumonia-associated mortality increases directly with age. The clinical impact of pneumonia in the elderly extends far beyond the time of the initial infection, in some cases extending up to a year afterwards, and is associated with significant mortality.^{31, 32} Some studies report that pneumonia rates for elderly patients in LTCF are 6-10 times that for community-dwelling elderly.³³ The presenting signs and symptoms of pneumonia may be muted in this age group, with tachypnea often the only clinical sign. Up to a third of patients will not exhibit fever or leukocytosis, at least initially.

· Community-Acquired Pneumonia (CAP) vs. Health Care-Associated Pneumonia (HCAP). As a general rule, as one goes from the community to the nursing home and hospital settings, "usual" community pathogens such as S. pneumoniae, H. influenzae, Moraxella catarrhalis, the "atypical" agents, and the respiratory viruses are replaced by gram-negative bacilli, Staphylococcus aureus (including MRSA, 20-40% of the time³⁴), and anaerobes in cases of aspiration associated with poor dentition. The distinction between nursing home and hospital-associated pneumonias has become blurred in recent years because of the back-and-forth traffic of patients between these two settings. This has led to the concept of "health care-associated pneumonia" (HCAP), meaning pneumonia that develops in patients who (a) were hospitalized during the preceding 90 days; (b) residents of nursing homes or extended-care facilities; (c) recipients of IV antibiotic therapy at home; (d) chemotherapy recipients; (e) individuals receiving wound care during the preceding 30 days; and (f) patients receiving long-term dialysis.^35,36,36a

· Community-Acquired Pneumonia. Major pathogens are Streptococcus pneumoniae, Haemophilus influenzae, Moraxella catarrhalis, Legionella pneumophila, Chlamydophilia pneumoniae, and respiratory viruses such as influenza and respiratory syncytial virus. S. aureus should be considered, along with S. pneumoniae, in those with post-influenza pneumonia, and P. aeruginosa should be considered in those with advanced COPD or bronchiectasis, especially if they have received outpatient antibiotics. Recent data has documented an increase in the incidence of legionellosis in the US since 2003, although most of this increase has occurred in those <65 years.³⁷

· CAP in the Elderly.

(1) Reduction in the incidence of invasive pneumococcal disease in the elderly. Since the introduction of the conjugated 7-valent pneumococcal vaccine for use in children, there has been a 38% decrease in the rate of invasive pneumoccal disease in elderly adults, presumably due to herd immunity.³⁸ Investigators from Omstead County, MN reported an 86% decrease in case-fatality rates for invasive pneumococcal disease for persons ≥65 years between 1995-1999 and 2001-2007.³⁹

(2) Pneumococcal Resistance to Penicillins and Cephalosporins. Resistance rates of S. pneumoniae to penicillin (PCN) have also changed, likely due to reduced nasopharyngeal carriage and transmission of antibiotic-resistant vaccine-type pneumococci from children to adults.^40-42 A study of 1647 pneumococcal isolates from 41 US medical centers indicated that while the prevalence of isolates with intermediate PCN resistance (MIC, 0.1-1.0 mcg/mL) increased from 12.7% to 17.9% from 1999-2000 to 2004-2005, PCN resistance (MIC ≥2 mcg/mL) decreased from 21.5% to 14.6%.⁴³ The vast majority of pneumococcal isolates remain susceptible to third-generation cephalosporins such as ceftriaxone or cefotaxime. The clinical significance of in vitro resistance to penicillins and the third generation cephlosporins has been assessed. In a thorough review of the literature, Peterson was able to find only a single report of documented microbiologic failure of parenteral penicillin-class antibiotics in the treatment of pneumococcal pneumonia in patients with or without bacteremia.⁴⁴ It turns out that most patients with pneumonia (but not meningitis) due to pneumococcal isolates with MICs of <4 mcg/mL may still be successfully treated with high dose penicillin (e.g., 3.2 million units IV q4h with normal renal function), oral amoxicillin (1 gram TID), or standard doses of ceftriaxone or cefotaxime. Unfortunately, many penicillin-resistant S. pneumoniae are also resistant to other antibiotics, including macrolides/azalides (erythromycin, clarithromycin and azithromycin), clindamycin, doxycycline, and trimethoprim-sulfamethoxazole. Currently 9.4% of pneumococcal isolates are resistant to clindamycin, 16.2% to tetracyclines, and 31.9% to trimethoprim-sulfa.⁴⁵

(2) Resistance of S. pneumoniae to Macrolides (clarithromycin) and Azalides (azithromycin). Macrolide resistance, currently defined as an MIC of ≥16 mcg/mL, is an increasing problem. Clinical failures are increasingly reported even where isolates exhibit lower-level (MIC of 1-8 mcg/mL) resistance. Currently about 30% of pneumococcal isolates in many parts of the United States are resistant to macrolides/azalides in vitro. Daneman et al⁴⁶ assessed the incidence of macrolide failures in a prospective population-based surveillance study conducted in two Canadian cities for the period between 2000 and 2004. They found 1696 episodes of pnemococcal bacteremia of which 60 (3.5%) were failures of macrolide therapy – that is, continued bacteremia despite outpatient macrolide therapy. Macrolide failure occurred even in those with low-level resistance (1 mcg/ml), and higher levels of resistance did not correlate with a higher likelihood of macrolide failure. It has been estimated that with a 25% rate of macrolide resistance, mortality due to inadequate therapy would occur in about 1 in every 100 patients with empirical macrolide monotherapy, with higher rates among the elderly and those with medical comorbidities.⁴⁷

(3) Resistance to Respiratory Fluoroquinolones (FQ). Although <1% of S. pneumoniae are resistant to the "respiratory" fluoroquinolones (e.g., levofloxacin, moxifloxacin, gemifloxacin) in most areas of the US, some states have reported rates in the range of 5%. Unfortunately, the reported rates of fluoroquinolone resistance may be somewhat misleading. Resistance most often results from mutation in the genes encoding the topoisomerase and DNA gyrase enzymes, namely parC and gyrA. Isolates with only the parC mutation will appear susceptible to the respiratory fluoroquinolones. However, once a parC mutation is present, gyrA mutations occur with a frequency in the range of 1 in 10⁵.⁴⁸ Since there are 10¹⁰ to 10¹² pneumococci in the lungs of patients with pneumococcal pneumonia, the likelihood of the second mutation, and thereby phenotypic resistance, is high. Currently, about 21% of S. pneumoniae in the United States already harbor a first-step resistance mutation, mostly parC.⁴⁹ What we may see in coming years, therefore, is a rise in FQ resistance.^50,51 Outside of the US, higher levels of FQ resistance are already observed. In Hong Kong, for example, > 13% of pneumococcal isolates are resistant to levofloxacin.⁵² The majority of levofloxacin-resistant isolates are also cross-resistant to the other FQ, and thus the entire FQ class is threatened. Fluoroquinolone resistance rates are increased among elderly in long-term care facilities.^{53, 54} Kupronis et al found that pneumococcal strains from LTCF patients were significantly more likely to be nonsusceptible to levofloxacin than strains from community-living elderly patients (5.7% vs. 0.4%, P < 0.001). Because of the emerging potential of FQ resistance, the CDC has recommended that beta lactam-based regimens be used in preference to FQ in the empiric treatment of CAP, with FQ being reserved for cases of beta lactam allergy or clinical failure.⁵⁵

(4) Two other problems with fluoroquinolone use: "collateral damage" and epidemic C. difficile. (a) Emerging pneumococcal resistance is not the only reason why we should limit the use of fluoroquinolones. Fluoroquinolones are active against a variety of gram-negative bacilli. Overuse of fluoroquinolones contributes increasing resistance among gram-negative pathogens including Pseudomonas aeruginosa.⁵⁶ While these agents "cover" the relevant CAP flora, they also cause "collateral damage" whereby their use leads to unintended resistance in flora irrelevant to the infection being treated. In addition to selecting for resistant gram-negative bacilli, fluoroquinolones may select for MRSA as well.⁵⁷ Of 150 episodes of MRSA colonization and 23 of MRSA infection, fluoroquinolones were the only antimicrobials that appeared to increase the risk of MRSA colonization (adjusted hazard ratio, AHR, 2.57; 95% CI 1.84-3.60, and AHR 2.49; 95% CI 1.02-6.07, respectively). Possible reasons for this include the disruption of the patient's complex microbiological flora, the selective inhibition of susceptible strains, and an increase in bacterial adhesion to surface fibronectin binding proteins after fluoroquinolone exposure. (b) An increase in the rate and severity of C. difficile – associated diarrhea has been linked to fluoroquinolone use in recent years (see C. difficile, below). The epidemic strains of C. difficile, which are resistant to fluoroquinolones, produce massive amounts (16-23 times higher than previous strains) of the toxins that cause diarrhea and colitis. The burgeoning use of fluoroquinolones in recent years may have selected for these previously uncommon strains, leading to C. difficile epidemics both in the U.S. and elsewhere.

· Treatment of Community-Acquired Pneumonia. Three general points regarding CAP therapy bear emphasis. (1) Antibiotics should be started as soon as possible after diagnosis. The prompt provision of antibiotics is now being used as a quality of care indicator by the Center for Medicare Medicaid Services and JCAHO. Initially, a 4-hour rule was imposed based on Medicare hospitalization data indicating an association between starting antibiotics within 4 hours of arrival and improved outcomes and reduced in-hospital mortality (severity-adjusted OR, 0.85; 95% CI 0.76-0.9), as well as shorter (0.4 days) mean length of stay.⁵⁸ The veracity of evidence behind the 4-hour rule has been questioned, although the principle of promptly starting antibiotics was affirmed.⁵⁹ One of the unfortunate consequences of the 4-hour rule was the overuse of antibiotics. Clinicians pressured to abide by this rule erroneously gave antibiotics prior to confirming the diagnosis of pneumonia.^{60, 61} Kanwar et al found that more patients in 2005 (28.5% vs. 20.6%, p = .04) had a hospital admission diagnosis of CAP without radiographic abnormalities than in 2003 when the 4-hour rule was established.⁶² The final diagnosis of CAP dropped to 58.9% in 2005 from 75.9% in 2003 (p < .001). Because of this issue, the time window time for administering the first dose of antibiotics has been extended to 6 hours. (2) Empiric initial antibiotic therapy should include coverage for "atypical" (Legionella pneumophila, Chlamydophylia pneumoniae) as well as "typical" (i.e., S. pneumoniae, H. influenzae, and M. catarrhalis) pathogens. This concept is supported by data from the National Pneumonia Project⁶³ and by previous reports.^64,65,65a and by more recent data.^65b In a retrospective review by Bratzler et al of 27,330 community-dwelling, immunocompetent Medicare patients age >65 years hospitalized in 1998-1999 and 2000-2001, the combination of a cephalosporin + a macrolide was associated with a 30% reduction in 14-day and 30-day mortality rates compared to a third-generation cephalosporin alone.⁶³ (3) Knowledge of antibiotics received by the patient in the three months prior is important to determine appropriate therapy patients with CAP.⁶⁶ Patients with recent penicillin, trimethoprim-sulfamethoxazole, and azithromycin use are at increased risk of harboring penicillin-resistant pneumococci. More than half of pneumococcal isolates recovered from patients with invasive disease who took azithromycin in the prior 3 months were macrolide-resistant. On the other hand, the absence of recent exposure does not necessarily predict macrolide-susceptibility. In one study, 55% of patients with macrolide-resistant pneumococcal bacteremia reported no antimicrobial drug exposure in the preceding 6 months.⁶⁷ Previous use of fluoroquinolones is a risk factor for fluoroquinolone-resistant pneumococci, as is residence in a nursing home.⁵⁴ Finally, patients with advanced COPD and a history of multiple previous antibiotic courses are at risk for pneumonia due to Pseudomonas aeruginosa.^68,35

· Guidelines for Empiric Therapy of CAP. Consensus guidelines of the Infectious Disease Society of America and the American Thoracic Society were published in 2007.⁶⁹ A simplified guideline, based on this publication and others, is outlined in the foregoing. Implicit in these recommendations are the principles of (a) using beta lactam-based regimens where possible to limit the overuse of fluoroquinolones for reasons cited above, (b) switching to oral therapy as soon as feasible to shorten length of stay (switch therapy); and (c) narrowing the spectrum of coverage if culture data allows; and (d) shortening the length of antibiotic therapy to lessen the risk of emerging resistance. For most patients admitted to the hospital: start with intravenous ceftriaxone or cefotaxime + azithromycin. If the patient is truly hypersensitive to penicillin (i.e., not just a rash, where third generation cephalosporin use would be permissible), then use a potent anti-pneumococcal FQ such as moxifloxacin or high-dose levofloxacin. Fluoroquinolone monotherapy should not be used for patients ill enough to require ICU admission, however.^65a Once the patient is clinically stable for 24 hours (systolic BP ³ 90 mm Hg, heart rate £ 100 beats/min, RR of £ 24 breaths/min, temp £ 37.2° C⁷⁰, then switching to oral antibiotics is appropriate.^{71, 72}. If reliable sputum culture data becomes available, then one would select a narrow-spectrum agent based on susceptibility results. For example, if an ampicillin-susceptible pneumococcus or H. influenzae is cultured from sputum and/or blood, then amoxicillin would be appropriate. Positive urine antigen tests for pneumococcus or Legionella may also enable narrowing of antibiotic coverage.⁷³ If no culture data is available to guide switch therapy, then one may use an empiric regimen such as high-dose amoxicillin + doxycycline or azithromycin to finish a total of 5-7 days of therapy. Fluoroquinolones such as moxifloxacin or levofloxacin should be reserved for those who are allergic to the above agents. Shorter courses of therapy, e.g., 5-7 days, may be as effective as traditional longer courses of 10-14 days, with less antibiotic exposure and less resistance pressure.⁶⁹ For outpatient therapy of CAP: use one of the empiric oral switch therapies listed above (i.e., amoxicillin +doxycycline or azithromycin, or a fluoroquinolone if the patient is allergic) as initial therapy.

· Treatment of Health Care-Associated Pneumonia (HCAP). The basic strategy for treating HCAP is to start with broad-spectrum therapy and then, very importantly, to de-escalate to narrower spectrum antibiotics once culture results becomes available. While pneumonia due to CAP flora can occur in healthcare settings, pathogens in HCAP most often include gram-negative bacilli, S. aureus including MRSA⁷⁴, and possibly anaerobes in cases of aspiration where the patient has poor dentition. Coliforms and other gram-negatives such as Pseudomonas aeruginosa account for approximately 40-60% of pneumonia in the nosocomial setting, and the majority of pneumonias acquired in the ICU. Since clinical outcome is tied to using effective antibiotics, initial empiric therapy should be broad-spectrum. Local patterns of antimicrobial resistance, the patient's past history of colonization (e.g., MRSA), and recent antibiotic exposure history should all factor into the choice of initial therapy. Sputum for C+S should be expeditiously obtained prior to starting antibiotics, if possible. Initial coverage should include antibiotics effective against gram-negative bacilli and S. aureus. Thus antibiotics with activity against gram-negatives and oxacillin-susceptible S. aureus such as cefepime (or Zosyn or imipenem if antibiotic-resistant gram-negatives are suspected; alternatively ceftazidime or aztreonam which lack activity against S. aureus) +/- vancomycin would be appropriate. For severely ill or ventilated patients, a second gram-negative agent such as ciprofloxacin may be added. Once culture and susceptibility results become available after 1-3 days, therapy should be de-escalated to narrower-spectrum agents.³⁶ This strategy of initial broad-spectrum therapy followed by de-escalation leads to a higher proportion of patients being treated with effective therapy from the outset. However, forgetting to de-escalate therapy needlessly exposes patients to broad-spectrum antibiotics and encourages the emergence of resistance. Provided that the patient is clinically improving after initial therapy, one should plan on continuing antibiotics for approximately 7 days. If the patient is not responding to the antibiotics, then invasive interventions such as bronchoscopy with bronchoalveolar lavage or protective specimen brush and quantitative cultures, along with studies to look for unexpected or opportunistic pathognes may be considered with attendant modification of antibiotic therapy if needed. If a patient with poor dentition develops pneumonia in the setting of gross aspiration, then anaerobic coverage should be provided. This can be accomplished either by adding metronidazole to a regimen such as vancomycin + cefepime (or ceftazidime or aztreonam), or substituting agents that cover both gram-negatives and anaerobes such as piperacillin-tazobactam or imipenem for the cefepime.

· The importance of oral hygiene in preventing pneumonia in the institutionalized elderly was recently reviewed.⁷⁵ Randomized controlled trials reveal positive effects of oral hygiene on preventing pneumonia and respiratory tract infection in hospitalized elderly persons and elderly nursing home residents, with absolute risk reductions of between 6.6% -11.7%, and numbers needed to treat to prevent infection from between 8.6-15.3 individuals. The authors estimate that approximately 1 in 10 cases of death from pneumonia in elderly nursing home residents may be prevented by improving oral hygiene. Quagliarello et al⁷⁶ identified inadequate oral hygiene (daily brushing of teeth and gums and regular dental care) as a modifiable risk factor in a prospective study of 613 nursing homes in New Haven (hazard ratio 1.60, 95% CI 1.06-2.35). The reason for this association remains speculative. Colonization with periodontal pathogens such as Prophyromonas gingivalis can degrade fibronectin on oral mucosal epithelium and enable colonization with potential respiratory pathogens such as gram-negative bacilli and S. aureus. In addition, stimulation of substance P by oral hygiene may facilitate swallowing function and reduce the risk of aspiration.

· Antipsychotic drug use and risk of pneumonia in the elderly. A nested case control analysis was recently reported of a cohort 22,944 elderly persons with at least one antipsychotic prescription in which 543 cases of hospital admission for pneumonia were identified.⁷⁷ Cases were matched with four randomly selected controls, and multivariate logistic regression analysis was used to estimate odds ratios. Current use of antipsychotics was associated with an almost 60% increase in the risk of pneumonia (AOR, 4.5; 95% CI, 2.8-7.3). Reasons proposed for this observation include drug-induced dyskinesia or spasm of the oropharyngeal musculature leading to aspiration, xerostomia leading to impaired oropharyngeal bolus transport, and sedation leading to swallowing problems.

· Pneumococcal vaccination. Pneumococcal disease kills more people in the United States each year than all other vaccine-preventable diseases combined. The indirect benefit of the pediatric conjugate pneumococcal vaccine in reducing rates of invasive pneumococcal disease among adults was discussed above. Otherwise healthy elderly persons should receive one dose of 23-valent capsular polysaccharide vaccine (PPV) at age 65. Those who previously received a dose of this vaccine due to underlying medical illness should have their second dose at age 65 or later. Repeating doses of PPV beyond two is not recommended. The data suggests that recipients of multiple booster doses exhibit immunologic hyporesponsiveness – that is, immune responses to repeated PPV doses that are actually less than to the initial PPV dose.⁷⁸ Newer conjugate pneumococcal vaccines for adults which stimulate T-cell dependent immune responses, similar to the current pediatric vaccine, are under investigation.⁷⁹

· Effectiveness of Pneumococcal Vaccination. Although the 23-valent pneumococcal capsular polysaccharide vaccine has never been shown to be effective in a randomized controlled trial, numerous case-control and epidemiologic studies support its use. Musher et al recorded the vaccination status of every patient for whom a culture yielded S. pneumoniaeduring a 4.5 year period.⁸⁰ They found that the rate of PPV vaccination was lower among patients with bacteremic pneumococcal pneumonia (39.7%) or any invasive pnemococcal disease (38.0%) than among patients with nonbacteremic pneumonia (57.6%). Fisman et al⁸¹ investigated the impact of prior PPV vaccination on in-hospital mortality and the probability of respiratory failure among hospitalized adults with CAP. Among 62,918 adults hospitalized at 109 community and teaching hospitals in the U.S. between 1999 and 2003, vaccine recipients were less likely to die of any cause during hospitalization than were those with no record of vaccination (adjusted OR, 0.50; 95% CI, 0.43-0.59) even after adjusting for the presence of co-morbid illnesses, age, smoking, influenza vaccination and under varying assumptions about missing vaccination data. Vaccination also lowered the risk of respiratory failure (adjusted OR, 0.67; 95% CI, 0.59-0.76) as well as other complications, and reduced the median length of stay by 2 days compared with non-vaccination (P<0.001). In a prospective cohort study conducted in Spain among community-dwelling elderly between 2002-2005, PPV was associated with significant reductions in the risk of hospitalization (hazard ratio, 0.74, 95% CI, 0.59-0.92), in the overall pneumonia rate (HR, 0.79; 95% CI, 0.64-0.98), provided significant protection against pneumococcal pneumonia (HR, 0.55; 95% CI, 0.34-0.88), and reduced risk of death due to pneumonia among vaccinated subjects (HR, 0.41, 95% CI, 0.23-0.72).⁸² Finally, investigators from Canada prospectively collected data on all adults admitted to 6 hospitals in an integrated health delivery system from 2000-2002.⁸³ The median age of 3415 evaluated patients was 75 years. Patients with CAP who had prior PPV vaccination had about a 40% lower rate of mortality or ICU admission compared with those not vaccinated.

Urinary Tract Infection in Elderly Patients

· Morbidity, Mortality, and Cost of Urinary Tract Infection. Urinary tract infection is the most common hospital-acquired infection, accounting for about 40% of all nosocomial infections,^84,85 affecting an estimated 8-900,000 patients per year in the U.S.⁸⁶ Up to 80% of nosocomial UTIs are associated with the use of urinary catheters, and these infections lead to increases in local and systemic morbidity, secondary bloodstream infection, death, a reservoir of drug-resistant microorganisms, and increased healthcare costs.^87,88 15-20% of short-term care patients in the U.S. receive urinary catheters during their hospital stay. Less than 1% of catheter-associated urinary tract infections lead to bloodstream infection,⁸⁹ but since urinary catheter use is so common, catheter-associated UTI is the most common source of nosocomial gram-negative bloodstream infection. As many as 15-17% of nosocomial bloodstream infections are secondary to UTIs.⁹⁰ In addition to the infectious complications of urinary catheters, important non-infectious complications include delayed discharge and lack of patient mobility.⁹¹ In the United States, the cost of diagnosing and treating catheter-associated urinary tract infection in university medical centers has been reported to range from $401 to $676 per episode.⁹² Data from the University of Michigan Health System suggest that the minimum cost of evaluating and treating a patient with urinary catheter-associated bacteremia is $2,836.⁹² Medicare's recent decision to decline reimbursement for the cost of treating preventable infections such as catheter-associated UTI provides a financial, in addition to medical, incentive to mitigate this problem.

· Important points regarding the prevention and treatment of urinary tract infection in the elderly include the following. (1) Catheter-associated UTI rates may be lowered by (a) avoiding catheter placement unless clinically indicated, and (b) by removing catheters as soon as feasible. (2) Avoid treating asymptomatic bacteriuria in the elderly. (3) Avoid treating pyuria and bacteriuria in patients with chronic indwelling urinary catheters unless clinically indicated.

· (1a) Approximately 30% of initial urinary catheterizations are unjustified, and one-third to one-half of continued catheterization days are unjustified.^92-94 Individuals with an indwelling urinary catheter develop bacteriuria at a rate of 3-10% per day, and the incidence of bacteriuria is universal in those who are chronically catheterized.⁹⁵ Studies have shown that physicians are often unaware that their patients have urinary catheters, which are commonly placed in the emergency room or operating room.^90,96 Appropriate indications for placement of urinary catheters include the following:⁸⁸

· Urinary retention

· Obstruction to the urinary tract distal to the bladder

· Close monitoring of urine output in critically ill patients

· Accurate measurement of urine output in an uncooperative patient (e.g., because of intoxication)

· Fluid challenge in patients with acute renal insufficiency

· Preoperative insertion for patients going directly to the operating room

· Comfort care in terminally ill patients

· Urinary incontinence that poses a risk to the patient (e.g., because of major skin breakdown or a nearby surgical site)

Inappropriate indications for placement or continued use of urinary catheters include the following:

· No longer needed for monitoring of urine output

· Unclear indication in patients for whom the catheter serves no useful purpose

· Urinary incontinence without significant skin breakdown

· Neurogenic bladder for which intermittent self-catheterization is possible

· Convenience of care

· Staff are too busy or forgot to remove catheter

As noted above, approximately 30% of initial urinary catheterizations are unjustified! Every clinician should carefully consider whether there is an appropriate indication before placing a urinary catheter, and review their continued need on a daily basis. Catheters are direct conduits for infection. The end must justify the risk.

· (1b) The second modifiable risk factor is limiting the duration of urinary catheterization. One-third to one-half of continued catheterization days are unjustified. A recent review of national data indicates that most hospitals in the U.S. do not have a system in place for keeping track of which patients are catheterized, and for monitoring catheter duration; nor do they employ measures of proven benefit such as scheduled reminders to remove catheters when no longer needed.⁹⁷ Several studies document success with interventions prompting clinicians to remove urinary catheters. Huang et al⁹⁸ reduced catheter-associated UTI in their intensive care units by prompting physicians to remove unnecessary catheters after 6 days of placement. Saint et al⁹⁹ reported a 26-41% relative decrease in duration of catheterization using a simple reminder to aid physicians in remembering that the patient had a urinary catheter. Apisarnthanarak et al⁸⁸ achieved a 73% reduction in the duration of urinary catheterization utilizing an intervention team that reviewed indications for continued catheterization, and discussed the advisability of removing the catheter with clinicians. With this intervention, the authors documented a dramatic reduction in catheter-associated UTI especially in their intensive care units, from 23.4 UTI/1000 catheter-days to 3.5 infections per 1000 catheter-days, P = 0.01.

· (2) Avoid treating asymptomatic bacteriuria in the elderly. Asymptomatic bacteriuria is defined as: (a) 2 consecutive voided urine specimens with isolation of the same bacterial strain in quantitative counts of ³ 10⁵cfu/ml in the absence of signs or symptoms of UTI in women; (b) a single, clean-catch voided urine specimen with 1 bacterial species isolated in a quantitative count of ³ 10⁵cfu/ml for men; or (c) a single catheterized urine specimen with 1 bacterial species isolated in a quantitative count of ³ 10²cfu/ml for either women or men.¹⁰⁰ It is estimated that about 10% of men and 20% of women ³ 65 y.o. are bacteriuric, and the prevalence increases with age for both sexes. The prevalence is higher among those in LTCF, in the range of 20-50% for non-catheterized patients. Patients with the greatest functional impairment including those with dementia and urinary and/or fecal incontinence are most likely to be bacteriuric. Patients with chronic indwelling catheters are virtually always bacteriuric. Antibiotic therapy for asymptomatic bacteriuria, even in the presence of pyuria, is not indicated. Prospective randomized trials have consistently shown that antibiotic therapy neither reduces the incidence of symptomatic infection nor improves survival in this setting. Moreover, treatment of asymptomatic bacteriuria does not lead to improvement in chronic genitourinary symptoms such as incontinence. Minimum criteria for antibiotic treatment of symptomatic UTI in patients without indwelling catheters includes fever (> 37.9C) with at least one of the following: new or worsening urgency, frequency, suprapubic pain, gross hematuria, CVA tenderness, or urinary incontinence.¹⁰¹ In the absence of these signs and symptoms, urine cultures should not be obtained.¹⁰²

· (3) Avoid treating bacteriuria and pyuria in patients with chronic indwelling catheters in the absence of systemic signs of infection attributable to the urinary tract. Approximately 10% of long-term care residents are catheterized primarily for incontinence or bladder outlet obstruction. Urinary tract infection usually follows the formation of biofilm on both the internal and external surface of the catheter. The chronically catheterized urinary tract always harbors bacteria, thus reducing the diagnostic value of urine cultures in these patients. Urine cultures should be obtained only as part of a work-up for systemic infection where all other causes of infection are considered first. Because bacteria causing infection may differ from those present in biofilm, the preferred specimen for culture is obtained from the sampling port of a newly-placed catheter immediately prior to the initiation of antibiotics.¹⁰³ Minimum criteria for the initiation of antibiotic therapy include the presence of any one of the following: fever, new costovertebral angle tenderness, rigors, or new onset of delirium.¹⁰¹ All too often, clinicians "chase" positive urine cultures, and prescribe unnecessary antibiotic treatment that selects for increasingly multi-resistant flora. In addition, unintended complications of such therapy include C. difficile-associated diarrhea and adverse drug reactions which may have devastating clinical consequences. The primary recommendation is to avoid culturing urine and prescribing antibiotics in the absence of a minimum set of symptoms or signs of urinary tract infection.

Tuberculosis.

· Current Epidemiology. The incidence of tuberculosis has been steadily declining over the past decade. New TB cases reported in the U.S. are at the lowest level in recorded history – 12,904 cases in 2008, or 4.2 cases/100,000.¹⁰⁴ Among the elderly the rate of newly diagnosed TB cases declined from 17.7 per 100,000 in 1993 to 6.4 per 100,000 in 2008. Currently, 19% of TB cases in the U.S. occur in those ≥65 years.

· Important aspects of the current guidelines include:

(1) emphasis on targeted tuberculin testing, focusing screening efforts to identify and treat those with latent tuberculosis infection (LTBI) – i.e., identifying those individuals infected with TB but without symptoms of active disease, who may serve as reservoirs for disease transmission should they reactivate;

(2) treatment of those with LTBI regardless of age;

(3) a uniform definition of skin test conversion regardless of age;

(4) revised treatment recommendations, including new regimens for LTBI therapy (9 months of INH as preferred therapy), and an emphasis on directly observed therapy both for treating those with active disease as well as for many persons with LTBI;

(5) permissible use, if available, of FDA-approved whole-blood interferon gamma interferon release assays such as the QuantiFERON®-TB or QuantiFERON-TB Gold test instead of tuberculin skin tests wherever skin tests might otherwise be used.^105,106

· Latent TB Infection (LTBI) vs. Active Disease. Among patients with intact immunity who become infected with TB, only about 10% will go on to develop active disease (c. 85% pulmonary TB, c.15% extrapulmonary or pulmonary + extrapulmonary). Half of these cases arise within 2 years of initial infection, and the other half at some later point in life. A number of conditions, in addition to advanced age, are associated with an increased risk of progression to active disease. Co-morbid conditions associated with an increased risk include silicosis (relative risk, RR, 30), diabetes mellitus (RR, 2.0-4.1), chronic renal failure (RR, 10.0-25.3), gastrectomy (RR, 2.0-5.0), and carcinoma of the head or neck (RR, 16). Immunosuppressive therapies used to treat patients with cancer or rheumatologic disease also increase the risk of reactivation. A recent study from Canada suggests that the adjusted RR of TB for disease-modifying rheumatologic drugs is 1.5 (95%CI, 1.1-1.9).¹⁰⁷

· Diagnosing tuberculosis infection. The two available methods of diagnosing latent TB infection include TST (formerly called PPD) as well as whole blood assays that measure interferon gamma release from monocytes in response to TB-specific antigens (e.g., QuantiFERON®-TB or QuantiFERON-TB Gold test). Due to lower cost and other factors, the TST is still most widely used. TST results should not be reported as "positive" or "negative", because there are three recognized cut-off points of 5, 10, and 15 mm, respectively, depending on the population tested. TST test results should always be reported as the number of millimeters of induration in the transverse diameter at 48-72 hours. Factors to consider when stratifying patients to one of the three cut-off points include (a) the risk of progressing from latent infection to active disease; and (b) the risk of exposure to tuberculosis. In the elderly population, the following would apply: 5 mm induration is considered positive for those at highest risk of progressing from latent infection to active disease, including (a) patients infected with TB as a result of recent exposure to a contagious case; (b) persons with fibrotic changes on CXR consistent with old healed TB (i.e., dense pulmonary nodules with or without visible calcification in the hilar area or upper lobes, sometimes with associated volume loss, and typically well-demarcated with sharp margins); and (c) patients who are heavily immunosuppressed (e.g., those receiving the equivalent of ³ 15 mg of prednisone/day for ³ 1 month). 10 mm induration is considered positive in persons who don't meet the preceding criteria but who have other risk factors either for exposure or for disease progression such as: (a) recent immigration (< 5 years) from countries with high TB prevalence; (b) residents of LTCF or other congregate residential facilities; and (c) persons with underlying co-morbid conditions conferring an increased risk of disease progression such as diabetes, silicosis, head and neck cancer, leukemia, Hodgkin's disease, end stage renal disease, etc. 15 mm induration is considered positive in persons with no known risk factors for TB – either for exposure or for progression to active disease.

· Clinical Approach to the Patient with a Positive TST. All persons with positive TST reactions are potential candidates for treatment regardless of age. A positive TST implies either latent infection (LTBI) or active disease. The first task is to determine if the patient has active disease, by assessing signs and symptoms (prolonged cough, hemoptysis, fever, night sweats, weight loss, etc.) and checking a chest X-ray. Those with LTBI are treated with isoniazid for 9 months (daily or twice weekly regimens). If isoniazid is not tolerated, then an alternative is to give rifampin for four months (daily). Patients with active tuberculosis are treated with multi-drug regimens. In general, the combination of isoniazid + rifampin + pyrazinamide + ethambutol is given during the initial two months (initial phase) while susceptibility testing on cultured isolates is performed. If the isolate is drug-sensitive, then ethambutol may be discontinued. After the initial phase, the course of further therapy is determined by the results of sputum culture at the two-month mark. If the two-month culture is negative, then patients are given the continuation phase of isoniazid + rifampin (daily or twice weekly) for 4 months, for a total of 6 months of therapy. On the other hand, if the sputum culture is still positive at 2 months and the patient has either cavitary pulmonary disease or is HIV positive, then isoniazid + rifampin is given for 7 months, for a total of 9 months of therapy. Because adherence to taking all of the prescribed doses is so critical to successful treatment, emphasis has been placed on using directly observed therapy (DOT) for the entire course of treatment. Clinical and laboratory monitoring. Because of the increased risk of hepatotoxicity from isoniazid, rifampin, and pyrazinamide in the elderly, baseline liver function tests should be obtained. Whether lab testing should be done only in response to symptoms of drug toxicity (e.g., anorexia, nausea, RUQ discomfort) or on a regular schedule has not been rigorously studied. However, periodic testing is advised for those with other risk factors such as regular alcohol use, or for those with abnormal LFT's at baseline. Isoniazid or rifampin should be withheld if the AST or ALT exceeds 3 times the upper limit of normal if associated with symptoms, and 5 times the upper limit of normal if the patient is asymptomatic. Protocols for re-implementing therapy should be followed once transaminases return to normal.

· Skin Test Conversion. The current CDC definition of TST conversion for persons of all ages is a ³ 10 mm increase in the diameter of induration compared with a result obtained within the previous two years. All persons who convert their TST should be evaluated for treatment, since conversion implies recent infection. Without treatment, the risk of developing active TB among elderly nursing home patients is in the range of 7.6 – 12.7%, with 90% developing active disease within a year of conversion.¹⁰⁸

· Boosting Reactions. In LTCF where regular (e.g., annual) testing of residents is recommended to detect converters, an initial two-step TST is recommended. If the first test is negative, then a second TST is performed a week or two later. If the second test is reactive, it is considered to be a boosting reaction reflective of prior TB infection, with the initial test being negative due to waning cutaneous hypersensitivity. The introduction of tuberculin antigen into the dermis with the first test, however, activates memory T cells such that cutaneous hypersensitivity is restored by the time of the second test. Patients with boosting reactions are considered TST reactive at baseline, with the following clinical implications: (1) being TST-positive, these patients should be evaluated and treated for LTBI; and (2) TST should not be performed again in the future since they are already positive at baseline. Those who are TST-negative after initial two-step testing, and who develop 10 mm or more of induration on subsequent (e.g., the following year) testing are considered to be skin test converters, and should be treated as such. It is important to use the same TST preparation year after year, in order to avoid false-positive tests. Gillenwater et al noted an increase in false-positive TST conversions when switching from Tubersol (Aventis Pasteur, Inc) to Aplisol (Parkdale Pharmaceuticals).¹⁰⁹

· Clinical Presentations of TB in the Elderly. A high index of suspicion for TB should be maintained at all times espec

«Back

News & Updates

Dr. Paul Kamitsuka Provides Lecture Notes on the Clinical Approach to Infections in the Elderly

Urinary Tract Infection in Elderly Patients

Features

Archive