Ten Clinical Situations in Long-Term Care for Which Antibiotics Are Often Prescribed but Rarely Necessary

Bacterial infections are prevalent in the long-term care (LTC) setting, and LTC facilities have particularly high rates of antibiotic use. Determining the appropriateness of antibiotic therapy represents a unique challenge for clinicians who care for LTC residents, as many residents have cognitive or expressive limitations that complicate efforts to assess symptoms. In addition, LTC residents can become septic quickly because they are often elderly and frail, making clinicians reluctant to delay treatment until culture results are available or the clinical situation fully presents itself.

Overuse of antibiotics in nursing home residents can lead to numerous complications, including antibiotic resistance, adverse reactions, multiple drug allergies, and secondary infections due to overgrowth of organisms like Clostridium difficile and Candida albicans.¹ Studies suggest that many LTC residents are colonized with drug-resistant bacteria.² Risk factors for harboring such bacteria include recent hospitalization, poor functional status, prior antimicrobial therapy (especially if recurrent), chronic pressure ulcers, and use of indwelling urinary catheters.² The growing number of bacterial strains demonstrating multidrug resistance, combined with the relative paucity of new agents in the antibiotic pipeline, has led antibiotic resistance to be declared a major public health crisis by several health-focused agencies, including the World Health Organization, the Institute of Medicine, the Infectious Diseases Society of America, and the Centers for Disease Control and Prevention.

To provide better guidance for clinicians working in LTC settings, we reviewed the literature and identified 10 clinical situations in LTC for which antibiotics are often prescribed but rarely necessary. We recommend alternative strategies that clinicians can suggest to nursing home staff in lieu of prescribing antibiotics (Table). The intent of these evidence-based recommendations is to refine prescribing practices for antibiotics, maximizing their efficacy and decreasing inappropriate use. When applying these recommendations in the practice setting, clinicians will have to evaluate each patient’s clinical situation and use their best judgment.

The prevalence of urinary tract infection (UTI) increases with age in men and women, and at least 30% of antibiotics prescribed in LTC are for treating these infections.³ We identified several scenarios involving UTIs in LTC residents for which antibiotic treatment is common but frequently unnecessary.

Positive Urine Culture in an Asymptomatic Patient

Asymptomatic bacteriuria refers to high quantities (>100,000 colony-forming units) of a uropathogen found in the urine of a patient who has no symptoms of a UTI. Urine cultures are positive for bacteriuria in approximately 25% to 50% of women and 15% to 40% of men who are not acutely ill and reside in LTC facilities.⁴ Studies show that asymptomatic bacteriuria does not increase mortality risk or chronic genitourinary symptoms, such as incontinence. Moreover, treating asymptomatic patients exposes them to the possible adverse effects of antibiotics without improving their prognosis.^5-9 Whereas an asymptomatic elderly resident with a positive urine culture does not require antimicrobial treatment,¹⁰ a resident with a positive urine culture who has clear symptoms arising from the urinary tract, such as dysuria, urinary frequency, or flank pain, should be treated for a UTI.

Urinalysis or Culture for Cloudy or Malodorous Urine

Nursing observations of foul-smelling or cloudy urine often lead to a urine culture being requested and occasionally to the empirical treatment of a suspected UTI.⁵ Although patients with symptomatic UTIs typically have cloudy or odorous urine, these signs are also observed in patients with noninfectious conditions, such as asymptomatic bacteriuria, dehydration, and crystallization after urine passage. Midthun and associates¹¹ found that one-third of nursing home residents who received a UTI diagnosis based solely on the odor of their urine tested negative for a UTI. Other studies have concluded that changes in urine appearance in the absence of symptoms do not indicate a need for urinalysis, culture, or antimicrobial treatment.^12,13 Instead, asymptomatic residents with malodorous urine should be managed with nursing interventions, such as scheduled toileting, perineal cleansing, and improved fluid intake, while concomitantly being monitored for new symptoms or a change in clinical status.¹⁴

For nursing home residents with a suspected UTI, urinalysis is used mostly for ruling out a UTI, rather than for ruling it in. Although the presence of nitrite is highly predictive for bacteriuria, asymptomatic bacteriuria is so common in this patient population that nitrite is only 20% specific for a UTI.¹⁵ Similarly, a dipstick test that is positive for leukocyte esterase is only 26% specific for a UTI because many noninfectious conditions can affect the bladder of older persons and cause pyuria.

A common practice in LTC facilities is to look for a UTI every time a resident falls, skips a meal, has an unusual episode of agitation, or falls asleep during Bingo. This contrasts starkly with a recommendation from a consensus panel of infectious disease specialists and geriatricians against treating UTIs in patients who have no localized signs or symptoms.^12,13 The best practice lies somewhere between these two poles.¹⁴ Although overtreatment of minor nonspecific symptoms is common, published diagnostic criteria lack adequate sensitivity.¹⁶

In the noncatheterized patient, nonspecific signs and symptoms and those not localized to the urinary tract are seldom due to a UTI.¹⁷ A nursing home study evaluated 399 cases in which a resident’s urine was cultured due to a suspected UTI.¹⁸ Overall, 37% of cultures were positive for bacteriuria plus pyuria and, in many cases, may have reflected asymptomatic bacteriuria. Investigators found that being female, the presence of dysuria, and a change in urine appearance were associated with bacteriuria, whereas fever and changes in behavior and mental status were not. Therefore, rather than reflexively deciding to “check the urine” when patients present with nonspecific signs and symptoms, physicians should consider other factors, such as poor sleep, constipation, mild dehydration, or drug-related adverse effects. In cases where clinicians decide that a urine culture is warranted, they should address other possible causes for the patient’s signs and symptoms while awaiting the results. If the results of the culture are positive and other conditions have been ruled out, clinicians should treat only those patients who remain symptomatic.

Management of respiratory infections in LTC is another area in which antibiotic use may be curtailed significantly. Antibiotic treatment does not improve many respiratory infections, including viral upper respiratory infections (URIs), many cases of bronchitis, and some cases of pneumonitis. Pediatricians have already succeeded in reducing the use of antibiotics to treat children with respiratory infections, with no corresponding increase in the incidence of serious infections. Decreasing antibiotic use is more challenging in the elderly population, which has higher prevalence rates of chronic respiratory pathology and serious bacterial infections.

Upper Respiratory Infections

URIs are common among older residents in LTC and are generally caused by viral pathogens, such as respiratory syncytial virus, rhinovirus, coronavirus, influenza, and adenovirus.¹⁹ A 2008 systematic literature review and meta-analysis of studies conducted among the general population confirmed that antibiotics are no more effective than placebo at improving URI symptoms and are sometimes associated with adverse effects.²⁰   One study concluded that URIs rarely cause serious complications and calculated that 4000 courses of antibiotics would have to be administered to prevent just one episode of post-URI pneumonia.²¹ The study noted, however, that individuals older than 65 years are more likely than younger adults to develop pneumonia following a URI and recommended monitoring older adults closely for additional symptoms that might require treatment.

Recommended supportive care measures for patients with a URI include rest, hydration, acetaminophen or nonsteroidal anti-inflammatory drugs to relieve fever and pain, and possibly steam inhalation to soothe throat irritation and nasal congestion. Although not specific to the elderly population, literature reviews of various nonprescription remedies for URI indicate that few have clinically significant benefits. In general, the efficacy of nasal saline irrigation,²² antihistamine monotherapy,²³ and antitussives²⁴ at relieving URI symptoms in adults has been inconsistent and limited. Antihistamine and decongestant combinations²³ and oral nasal decongestants alone²⁵ have demonstrated modest efficacy at providing short-term relief of some symptoms in adults. Considering the frail state of many elderly LTC residents, it is likely that the risk of adverse effects from over-the-counter URI drugs outweighs any modest benefits that might be realized. Not enough data are available from high-quality trials to support use of alternative medicines, such as zinc lozenges and echinacea.²⁴ Despite the common belief that vitamin C fights colds, studies have not shown vitamin C to reduce the duration of colds in symptomatic patients.²⁵

Bronchitis Absent of Chronic Obstructive Pulmonary Disease

Acute bronchitis is typically viral in origin. It often starts with rhinitis, with symptoms progressing to a persistent cough within a few days of onset. Abnormal vital signs (eg, fever, tachypnea, or tachycardia) and signs of consolidation or rales on physical examination suggest possible pneumonia and should prompt chest radiography.^26,27 Therapies are generally provided to relieve symptoms and may include analgesics, antipyretics, antitussives, bronchodilators, and expectorants. Studies show that patients whose bronchitis is associated with wheezing or a bothersome cough benefit from using inhaled bronchodilators.^28-31

Clinical trial data suggest that antibiotics may reduce the duration of bronchitis symptoms modestly at best; thus, antimicrobial agents are not recommended in most acute cases. Two meta-analyses of clinical trials comparing antibiotic treatment with placebo for patients with acute bronchitis found that antibiotic agents were only slightly better than placebo at relieving symptoms and that any improvement was minor.^28,32 The meta-analysis by Bent and associates,³² which included data from eight placebo-controlled trials, found that acute bronchitis symptoms in patients who received erythromycin, doxycycline, or trimethoprim/sulfamethoxazole decreased a fraction each day. Although the degree of improvement was statistically significant, the authors concluded it was too trivial to outweigh the risk of drug-related adverse effects.

An exception to recommendations against using antibiotics to treat acute bronchitis is when the patient has chronic obstructive pulmonary disease (COPD). Acute bronchitis often exacerbates COPD symptoms, and antibiotics have been found to reduce rates of mortality and treatment failure and to decrease sputum purulence.^33,34 Ram and associates³⁴ determined that eight patients needed to be treated to prevent one treatment failure. High-level evidence suggests that a 5-day course of antibiotics is adequate for most patients and is no less effective than longer regimens.³⁵

Suspected or Proven Influenza Without a Secondary Infection

Influenza ranks among the most commonly reported infectious disease outbreaks in nursing homes.³⁶ Because of its epidemic nature, influenza is frequently diagnosed based on clinical and epidemiologic findings during an outbreak. Influenza is caused by a viral pathogen; thus, treatment with antiviral agents may be appropriate, and antibacterial agents are only indicated if the patient has a secondary bacterial infection. When influenza is suspected, nasopharyngeal wash or swabs from the throat and nasopharynx should be taken and tested for influenza and other viruses, such as respiratory syncytial virus.^31,36,37

The most effective strategy for managing influenza is to institute an aggressive program that calls for immunizing residents and staff annually. If an influenza outbreak does occur, administering antiviral agents early in the course of the outbreak can help limit the spread of the virus.^36,38 Treating patients with antiviral medications promptly after the onset of illness may also reduce the severity and duration of uncomplicated influenza infections.³⁹

Respiratory Symptoms in a Terminal Patient With Dementia

Residents of nursing homes or assisted living facilities who have advanced dementia typically experience recurrent infections and frequent febrile episodes for which they often receive antibiotics.^40,41 Despite the frequent use of antibiotics to treat terminal patients with dementia, pneumonia is a leading cause of death in this patient population. Multiple studies have shown that antibiotics fail to prolong the life of dementia patients with pneumonia and are frequently associated with administration-related discomfort and adverse effects.^40-46 Although some physicians prescribe antibiotics in an effort to relieve pneumonia symptoms, several treatments other than antibiotics have been found to be far more palliative for dementia patients with pneumonia. Relief measures include analgesics for pain, anticholinergics (atropine or scopolamine) for respiratory mucous secretions, anxiolytics and neuroleptics for agitation, oxygen and narcotics for dyspnea, and ice chips and oral fluids for dry mouth.^47,48

Skin wounds are common in LTC and have many causes, including injury, surgical incision, immobility, incontinence, chronic disease, and contagion. Recognizing criteria of infection in skin wounds is difficult because the signs of chronic inflammation are similar to those of overt infection. Bacterial infection is a primary concern when managing skin wounds.⁴⁹ Uncontrolled infection can lead to discomfort, rapidly spreading tissue necrosis, and even death.⁵⁰ As a result, it is tempting to administer systemic antibiotics to residents with skin wounds, even without signs of deep infection. Optimal management can minimize the risk of bacterial infection, however, and several studies and guidelines provide alternative strategies to antibiotics for managing soft-tissue infections,⁵¹ pressure wounds,^52-55 and small abscesses.⁵⁵ They also offer guidance on determining when antibiotics are indicated.^50,56,57 

Skin Wounds Without Cellulitis, Sepsis, or Osteomyelitis

When managing skin wounds, providers should take care to determine whether a wound is actively infected before administering systemic antibiotics. Studies have demonstrated the value of using subtle signs to facilitate identification of wound infections^49,58; however, it has also been shown that signs of infection differ between wound types.⁵⁹ In 2009, Woo and Sibbald⁵⁶ compared quantitative swab cultures to the presence or absence of potential clinical signs of wound infection based on the existing assessment model using the mnemonic devices NERDS to determine  colonization levels (nonhealing, increased exudate, red friable granulation, debris, and smell) and STONEES to determine infection (increased size, increased temperature, Os [probes to bone], new areas of breakdown, edema/erythema, increased exudate, and smell). The results of this study seem to confirm that no one clinical sign of wound infection may be sufficient to help clinicians decide whether a topical or systemic antibiotic should be prescribed; however, the results suggest that the presence of three potential signs of infection in the wound can serve as good indicators of the extent of bacterial colonization, with a 73.3% sensitivity and 80.5% specificity for scant or light bacterial growth and 90% sensitivity and 69.4% specificity for moderate and heavy bacterial growth.⁵⁶ In a European study, cellulitis, malodor, pain, and delayed healing or deterioration of the wound (wound breakdown) were signs of infection that were common in all wound types.⁶⁰ Based on the aforementioned studies, close observation of the wound over time is imperative so that subtle changes can be identified, ensuring signs of infection are not missed.

For superficially infected wounds, optimal management requires conducting a thorough assessment and debridement of the wound. Use of topical dressings and agents, such as cadexomer iodine, antibiotic creams, and nanocrystalline silver agents, can also be considered.^52,61 Pressure ulcers should initially be cleansed at each dressing change, applying minimal mechanical force. The wound should be washed with normal saline or water at room temperature; skin cleansers and antiseptic agents should be avoided.⁵³ It is important to remove any slough and necrotic eschar, which serve as physical barriers to healing, support bacterial growth, and may mask signs of local infection. After debridement, use of a topical antimicrobial agent such as cadexomer iodine or silver-containing dressings should be considered, as these can restore bacterial balance to the wound.^62-64

When managing wounds, it is important to choose a dressing that keeps the wound environment moist but not saturated. This protects surrounding healthy tissue, which is typically torn when removing a dried-out dressing.⁵⁴ Absorptive dressings, hydrophilic foam, alginates, or saline-impregnated gauze are indicated for wounds that produce significant amounts of exudate, and the dressing should be changed daily.⁵⁷

Swabbing and culturing a skin wound is rarely recommended because a positive bacterial culture is more likely to indicate surface colonization rather than active infection. Thus, skin infection is usually a clinical diagnosis; however, when a wound stops healing or shows subtle signs of deterioration (ie, persistent purulent drainage), an appropriately deep culture from the wound bed should be obtained.⁵⁵

Small Localized Abscess Without Significant Cellulitis

Recent guidelines from the Infectious Diseases Society of America state that most abscesses can be effectively treated with incision and drainage alone, although cellulitis generally requires systemic antibiotics.⁶⁵ This is true even for abscesses caused by methicillin-resistant Staphylococcus aureus (MRSA).⁵¹ It is generally thought that systemic antibiotics are unnecessary for abscesses measuring 5 cm or less in diameter.⁶⁵ The role of antibiotics is less clear when treating MRSA-caused abscesses in patients who are immunocompromised or who have multiple comorbid conditions.⁵¹ Some studies suggest, however, that incision and drainage alone may be sufficient care even for these patients.^66,67

Decubitus Ulcer in a Terminal Patient

In patients with end-stage illness, decubitus ulcers are difficult to avoid and may represent a visual biomarker of their terminal status.⁶⁷ Treatment options are limited, and goals of care should focus on palliation and controlling pain and odor.^67,68 Managing wound odor is important and rarely requires systemic antibiotic therapy.^68-71 For superficial wounds, tissue debridement followed by topical metronidazole gel or silver sulfadiazine can often help control odor.^68,69,71 For
venous ulcers, the antimicrobial agent cadexomer iodine may decrease bacterial counts and odor. Data from a handful of studies suggest that applying yogurt or buttermilk to a clean wound lowers the pH of the wound, thereby limiting bacterial proliferation and odor.⁷² Generally, systemic metronidazole should only be administered for malodorous deep tissue infections.

In the majority of cases covered by these 10 scenarios, effective management of symptoms can be achieved with therapeutic options other than systemic antibiotics. Given how common suspected UTIs, respiratory symptoms, and skin wounds are in LTC, identifying situations where antibiotics are not indicated should significantly decrease their use, thereby reducing the risk of adverse outcomes, including antibiotic resistance. The clinical application of the recommendations outlined in this article should be individualized to each situation and will require concomitant education of nursing staff and families.

Authors: Christine Khandelwal, DO • Christine Lathren, MD, MPH • Philip Sloane, MD, MPH

Affiliations:
This research was conducted by the University of North Carolina at Chapel Hill under contract no. 290200710014I to the Agency for Healthcare Research and Quality, Rockville, MD. The authors of this article are responsible for its content. No statement may be construed as the official position of the Agency for Healthcare Research and Quality of the US Department of Health and Human Services.

Dr. Sloane is a recipient of Pioneer Award #PIO-04-1004 from the Alzheimer’s Association.

Disclosures:
The authors report no relevant financial relationships.