Criteria for Urinary Tract Infection in the Elderly: Variables That Challenge Nursing Assessment

Susan J. Midthun


Urol Nurs. 2004;24(3) 

In This Article

Criteria for Urinary Tract Infections


Colony Counts. The hallmark of a UTI has been the presence of a single microorganism of ≥ 105 colony forming units (cfus) per ml in a clean-catch or midstream urine specimen, with lower numbers usually indicating contamination (Kass, 1957). Studies of the general population have suggested lower colony counts for reasons which include: an early stage of infection, recent use of antibiotics, cleaning the perineum with an antiseptic, frequency of urination, urinary tract obstruction, state of hydration (including the time of day the specimen was collected), urine pH or specific gravity, or antibiotic therapy (Kass, 1957; Kunin, White, & Hua Hua, 1993; Pollack, 1983; Strand, Bryant, & Sutton, 1985).

In studies of UTIs in the elderly, there is a lack of agreement with research findings. One study recommended increasing the significant level of bacteriuria to 106 cfu/ml due to the spontaneous recovery of elderly individuals with 105 cfu/ml and the increased likelihood of contamination (Nordenstam, Sundh, Lincoln, Svanborg, & Eden, 1989). Conversely, another study found lower colony counts acceptable for elderly men (Lipsky, Ireton, Fihn, Hackett, & Berger, 1987). The third article did not dismiss the possibility of lower counts diagnostic for women as well with further research needed in this area (Gleckman, 1992).

Polymicrobial Bacteriuria. Contamination may be the most frequent cause of multiple microorganisms in the elderly (Bartlett & Treiber, 1984; Nickel & Pidutti, 1992). Nevertheless, though it is uncommon in the younger population, 25% to 33% of bacteriuria in LTCFs may be polymicrobic (Baldassarre & Kaye, 1991; Nicolle, Bjornson, Harding, & MacDonell, 1983; Nicolle, 1992). Reasons for multiple microorganisms, other than contamination, include fistulas, urinary retention, infected stones, or catheters (Nickel & Pidutti, 1992). In one study, of the 98% of catheter urine that contained high concentrations of bacteria, 78% of the cases of bacteriuria were polymicrobial (Warren, Tenney, Hoopes, Muncie, & Anthony, 1982). Polymicrobic bacteriuria has been associated with chronic infections and a higher mortality rate in both the elderly and the general population (Gross, Flower, & Barden, 1976; Mulholland & Weisman, 1987; Rudman, Hon tanosas, Cohen, & Mattson, 1988; Windsor, 1983). However, at present, polymicrobic bacteriuria has not been determined to be a risk factor for the elderly.

Absence of Symptoms. Asymptomatic bacteriuria (ASB) in the elderly is considered a benign and transient condition that does not require antibiotic treatment (Abrutyn et al., 1994; Eberle, Winsemius, & Garibaldi, 1993). This conclusion is based on studies showing that antibiotic treatment of ASB does not decrease morbidity or mortality in the elderly, is not effective in maintaining a sterile urine, causes side effects, and increases the incidence of reinfection and the number of drug-resistant strains of bacteria (Abrutyn et al., 1994, 1996; Mims, Norman, Yamamura, & Yoshikawa, 1990; Ni colle, Mayhew, & Bryan, 1987). Still, there are several precautions related to the recommendation of no antibiotic treatment for ASB in the elderly as shown in Table 2 (Hochreiter & Bushman, 1999; Nicolle, 1994; Nordenstam et al., 1989; Riehmann, 1998; Stamm, 1983; Zhanel, Harding, & Guay, 1990). It should be noted that the authors cited above do not share consensus with the parameters as outlined in Table 2 .


Pyuria indicates a host response to infecting bacteria and has been associated with the presence of both symptomatic and asymptomatic UTIs in the elderly (Aguirre-Avalos, Zavala-Silva, Diaz-Nava, Amaya-Tapia, & Aguilar-Benavides, 1999; Gleckman, 1992).

However, there are many reasons for pyuria in the elderly other than a UTI. Some researchers have found pyuria, both with ASB and without bacteriuria, to be so common in the elderly that they regard it as having questionable value, both in UTI detection and as an indication for therapy (Baldassarre & Kaye, 1991; Ouslander, Shapira, Schnelle, & Fingold, 1996; Reid & Nicolle, 1999; Rodgers et al., 1991). The issue of whether pyuria is diagnostic of a UTI has yet to be resolved. Jackson et al. (1992) and Garner et al. (1988) included pyuria as an indication of a UTI in institutionalized elderly. However, McGeer et al. (1991) rejected pyuria as being a reliable predictor of bacteriuria or symptomatic infection in this population.

Time of Laboratory Analysis. In the LTCF, laboratory facilities may not be readily available, causing a delay in specimen analysis (Bentley et al., 2001). Though not specific to the geriatric population, time before analysis has been discussed as a factor in assessing pyuria (Kierkegaard, Feldt-Rasmussen, HF8rder, Anderson, & JF8rgensen, 1980; Triger & Smith, 1966). Alkaline urine, such as is caused by urease-producing pathogens, may, over time, lyse cells and lower WBC counts of specimens containing these bacteria (Gnarpe & Edebo, 1965; Nicolle, 2001; Rodgers et al., 1991). However, Anderson, Johnson, and Trombley (1994) found that only urines with pH values of 9 or greater stored for more than 4 hours were at risk for decreased WBC counts due to cell lysis.


Even when UTI symptoms are present in the LTCF resident, they must be recognizable symptoms to justify further analysis. The major problem confronting the nurse is addressing what changes in the resident's condition indicate the need for further evaluation of a UTI.

New or Increased Urgency, Frequency, Dysuria. Classic signs of a UTI in younger individuals (urgency, frequency, and dysuria) occur in elderly patients as well (Baldassarre & Kaye, 1991; Beier, 1999; Fune, Shua-Haim, Ross, & Frank, 1998; Jackson et al., 1992; Johnson, 1991; Nicolle, 1994; Schaeffer, 1988; Whippo & Crea son, 1989; Yoshikawa, 1993). However, catheterized elderly with a UTI rarely express burning, and frequency and urgency do not occur (Yoshikawa & Norman, 1996). In the noncatheterized elderly, these symptoms may also be absent, masked, or difficult to assess, especially among the cognitively impaired (Berman, Hogan, & Fox, 1987; Colling, McCreedy, & Owen, 1994; Esposito et al., 1980; MacLennan, 1998; Nicolle, 1993). Nursing implications of the absence of these typical symptoms include a possible delay in diagnosis or a misdiagnosis resulting in poor outcomes in the frail elderly, who deteriorate more rapidly from infection than the younger population (Barkham, Martin, & Eykyn, 1996; Setia, Servanti, & Lorenz, 1984).

In addition, complaints of urgency, frequency, and dysuria can be common and chronic in the elderly without bacteriuria (Baldassarre & Kaye, 1991; Boscia et al., 1986; Nicolle, 1993). Further, although antibiotic therapy effectively sterilized the urine of asymptomatic elderly women, it did not stop genitourinary symptoms from developing (Abrutyn et al., 1994; Nicolle et al., 1987). Therefore, the presence of frequency, urgency, and dysuria, even with the occurrence of bacteriuria, requires careful interpretation as these symptoms may not be due to a UTI (McMurdo & Gillespie, 2000).

Change in Character of Urine. Cloudy, bloody, or malodorous urine was found in over 85% of symptomatic UTIs in one LTCF study (Jackson et al., 1992). Conversely, other studies that further delineated these symptoms found them less predictive of a UTI.

Cloudy. Nursing assessment of the clarity of urine requires that the urine be evaluated within 1 hour of voiding in a clear plastic or glass container (Schumann & Greenberg, 1979; Strasinger, 1985). Clarity is classified as clear, which is associated with a lack of bacteriuria; or hazy, cloudy, milky, or turbid, which is associated with bacteriuria and pyuria. Researchers have found bacteriuria in crystal clear urine, however (Pouchot, Launay, Cahen, Boussougant, & Vince nux, 1990; Tremblay & Labbé, 1994). Further, cloudiness occurs in normal urine due to epithelial cells, mucus, some types of crystals and casts, and talcum powder (Schumann & Greenberg, 1979; Strasinger, 1985; Urinalysis Today, 1986). In one study of the general population, 60% of cloudy urines were due to bacteriuria and 30% were attributed to amorphous (nonpathological) material (Schumann & Greenberg, 1979). Cloudy urine, alone or in combination with urine dipstick results, was helpful in determining the need for further laboratory analysis in one study of the elderly (Flanagan, Rooney, Davies, & Stout, 1989). However, Loeb et al. (2001) stated cloudy urine does not indicate the need for antibiotics.

Bloody. Numerous studies have cited reasons for hematuria other than a UTI (Mariani et al., 1989; Misdraji & Nguyen, 1996; Nicolle et al., 1993; Restrepro & Carey, 1989; Strasinger & DiLorenzo, 2001). In a study of LTCF residents, Nicolle et al. (1993) found that most episodes of hematuria in the LTCF were not due to a UTI. Common pathophysiologic conditions and medications associated with hematuria are summarized in Table 3 .

Malodorous. One study of LTCF residents found that urine odor or absence of odor misled the diagnosis of bacteriuria in over one-third of all cases (Midthun, Paur, & Lindseth, 2000). Further, because bacteriuria is common in nursing homes and is considered transient, bacteriuric odor as indicative of an infection has been questioned. In fact, in several studies foul-smelling urine was not considered a valid indicator of a UTI even with bacteriuria (Breiten bucher, 1990; Loeb et al., 2001). Nicolle (1993) stated that, though malodorous urine may be caused by bacteriuria, improved toileting may manage the problem better than antibiotic therapy.

Elevated Temperature. The elderly may require more time to demonstrate a fever, may show no increase or an increase of ≤2.4° F above an individual's baseline temperature, or may even be hypothermic in the presence of infection (Barkham et al., 1996; Castle, Norman, Yeh, Miller, & Yoshikawa, 1991; Colling et al., 1994; McAlpine, Martin, Lennox, & Roberts, 1986; Yoshikawa & Norman, 1996). The lack of fever can delay diagnosis and treatment, leading to poorer outcomes (Castle et al., 1991). Elderly individuals at increased risk for a muted or absent fever response are reported in Table 4 (Fune et al., 1998; Gleckman & Hibert, 1982; Keating, Klimek, Levine, & Kiernan, 1984; Kreger, Craven, & McCabe, 1980; Norman & Yoshikawa, 1990). Table 5 outlines several external factors affecting temperature variability in the elderly and nursing implications (Castle et al., 1991; Darowski, Najim, Weinberg, & Guz, 1991; Ellsworth, Witt, Dugdale, & Oliver, 1999; Esposito et al., 1980; Gleckman & Hibert, 1982; Katz, Beam, Brand, & Boyce, 1990; Meyers et al., 1989; Moe, Prinz, Vitiello, Marks, & Larsen, 1991; Norman & Yoshikawa, 1990; Yoshikawa & Norman, 1996).

Studies have also found that the elderly do not lack a febrile response; that an elevated temperature was the most common initial symptom and a marker for a serious infection and the most important clinical indicator for antibiotic treatment (Alessi & Harker, 1998; Berman et al., 1987; Katz et al., 1990; Keating et al., 1984; Yoshikawa & Norman, 1996). However, other studies found that some fevers resolved without treatment and that administering antibiotics without a clear indication of infection did not improve outcomes in the geriatric population (Finnegan, Austin, & Cape, 1985, Warren, Damron, Tenny, Hoopes, & Muncie, 1987).

An obvious difficulty in assessing fever is that an elevated temperature is not specific for a UTI or even an infection (Finnegan et al., 1985). Due to the frequent occurrence of bacteriuria in this population, assessing all fevers with bacteriuria as UTIs may be problematic (Warren et al., 1987). One study found that less than 10% of all elevated temperatures in the LTCF elderly were due to a UTI (Orr et al., 1996). In contrast, other studies have suggested that a greater percentage of febrile infections in LTCFs are caused by UTIs (Alessi & Harker, 1998; Jackson et al., 1992; Jacobson & Strausbaugh, 1990).

In the case of bacteremia, a very serious illness in the elderly that is most often caused by a UTI, the case for fever is again not consistent. In some studies, fever was one of the most common symptoms, often present in > 90% of cases (Chassagne et al., 1996; Esposito et al., 1980; Meyers et al., 1989; Richardson & Hricz, 1995; Rudman et al., 1988). Conversely, other research found an absence of fever or even hypothermia in bacteremic elderly (Ackermann & Monroe, 1996; Gleckman & Hibert, 1982; Kreger et al., 1980: Whitelaw, Rayner, & Willcox, 1992). Again, the lack of this symptom might delay diagnosis. This may be the reason for the high mortality rate for elderly with this condition within 24 hours of hospital admission (Whitelaw et al., 1992).

Chills. Chills are related to temperature and are a part of the body's behavioral and physiologic thermoregulation mechanism. In the elderly, both these areas may be impaired (Norman, Grahn, & Yoshikawa, 1985). Chills, however, have been found in bacteremic elderly (Chassagne et al., 1996; Fontanerosa, Kaeberlein, Gerson, & Thomson, 1992; Leibovici, Greenshtain, Cohen, Mor, & Wysenbeek, 1991; Rudman et al., 1988). Although UTIs are the major source of bacteremia in the elderly, there are other causes as well. Therefore, chills are important for assessing infection, even though they are not always specific for a UTI.

Pain. Studies are not in agreement regarding acute pain perception by the elderly (Harkins, 1996). Further, among the frail, demented, and very old, no studies were found. Pain from chronic diseases or other acute problems may mask or mislead diagnosis of acute diseases, or pain may not be specific for a particular disease (Norman & Toledo, 1992). Boscia et al. (1986) found that subjects reported no difference in UTI symptoms, such as suprapubic and flank pain, regardless of presence of bacteriuria. Whippo and Creason (1989) found low back pain common in LTCF residents and concluded that it could be due to factors other than a UTI such as arthritis or immobility. Gleckman (1992) found that abdominal pain caused by a UTI was initially diagnosed as a gastrointestinal problem. Medications may mask pain symptoms as well. Sociocultural differences and psychological factors may increase or decrease the likelihood that people will report pain. The cognitively impaired elderly may not be able to communicate the presence of pain. In addition, they may under-report symptoms or simply not understand a change in their condition (Beier, 1999; Irvine, Van Buren, & Crossley, 1984; Norman & Toledo, 1992). Agitation, irritability, restlessness, decreased appetite, in creased confusion, or falls may indicate pain in these individuals.

Despite these limitations, the elderly present with suprapubic, flank, or costovertebral angle pain as evidence of UTIs (Fune et al., 1998; Nicolle, 1994). In their 3-year study of nursing home infections, Jackson et al. (1992) found suprapubic tenderness in 18% of residents with symptomatic UTIs, though flank pain was present in only 4%, and costovertebral angle tenderness in less than 1%. Other studies have found suprapubic, costovertebral angle, or flank pain absent or infrequent in acutely ill elderly with UTIs (Berman et al., 1987; Esposito et al., 1980; Yoshikawa, 1993). Gleckman et al. (1982) stated: "An attempt to elicit costovertebral angle tenderness was often omitted suggesting that house staff considered this maneuver of no diagnostic value or difficult to assess in elderly acutely ill patients" (p. 552). Jackson et al. (1992) agreed that assessing pain related to infection in the elderly was difficult and required the expertise of a skilled health care professional.

Incontinence. Incontinence may be directly caused by a UTI or indirectly caused by the altered mental state that occurs with this infection (Karafin & Coll, 1987; Windsor, 1983). However, incontinence, which may be chronic in as much as 50% of the LTCF population, is commonly caused by other conditions (Karafin & Coll, 1987; Rowe et al., 1989). Not only does this make evaluation difficult, but the significance of bacteriuria in this incontinent population is also not clear (Berman et al., 1987; Ouslander, Kane, & Abrass, 1982; Rowe et al., 1989). Chronic incontinence among LTCF residents was not affected by urine sterilization (Ouslander et al., 1995).

Incontinence is not only a symptom but also a risk factor for a UTI (Molander, Arvidsson, Milsom, & Sandberg, 2000; Ouslander et al., 1982; Richardson & Hricz, 1995). In the case of LTCF residents with chronic incontinence, it may be more beneficial to manage the incontinence to control the bacteriuria instead of treating the bacteriuria with antibiotics to control the incontinence.

Worsening of mental or functional status. A decline in mental or functional status, as a symptom of a UTI, may be seen in the elderly. Several geriatric studies have identified and further defined these symptoms as evidence of a UTI (see Table 6 ) (Beier, 1999; Bentley et al., 2001; Esposito et al., 1980; Norman & Yoshikawa, 1990; Windsor, 1983; Yoshikawa, 1984a; Yoshikawa & Norman, 1996).

Confusion, lethargy, agitation. Toxicity from infection can cause an altered mental status (Gleckman et al., 1982). Yet, deterioration of mental status in a population with a high degree of cognitive impairment makes judging this indication of a UTI difficult (Norman & Toledo, 1992; Richardson & Hricz, 1995). Even so, research has shown that an altered mental state was the second most common indicator of bacteremia in the elderly (Meyers et al., 1989; Whitelaw et al., 1992; Windsor, 1983). In an analysis of several studies of bacteremia in the elderly, the majority of which were genitourinary in origin, lethargy was the second most common indicator of bacteremia and confusion the third (Richardson, 1993).

Falls. Though a fall experienced by an elderly patient is a vague symptom of a UTI, it has been used to indicate an infection (Beier, 1999; Berman et al., 1987; Richardson, 1993; Windsor, 1983). Though not specific for a UTI, if a fall indicates a change in the status of the resident such as increasing weakness, it should be used in evaluating the total clinical picture.

Appearance of the Patient. This is another vague symptom and illustrates, not only the distinctive nature of a UTI in an elderly individual, but the unique environment in which the assessment takes place. In the LTCF, where the nurse often does not see each resident on a daily basis, it is not uncommon for a family member or nonmedical staff member familiar with the person to report to the nurse that the resident is somehow different. This observation should not be dismissed, especially if the resident is cognitively impaired. A decline in general status has been described as an indication and sometimes the only indication of a UTI in the elderly (Fune et al., 1998; Katz, 1993). This symptom may signal the subtle physical, mental, or functional changes that are present but difficult to describe in many elderly. Moreover, it also illustrates the importance of the assessment by the LTCF nurse who knows the resident and is best able to identify the change. But, with the number of elderly bacteriuric individuals who do not benefit from antibiotic treatment, it is important that the nurse only use this symptom as a starting point for further evaluation.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.