Babies and infants

UTI is a problem in all age groups, although its prevalence varies markedly. In infants up to the age of 6 months, symptomatic UTI has a prevalence of about two cases per 1000 and is much more common in boys than in girls. In addition to these cases, asymptomatic UTI is much more common than this, occurring in around 2% of boys in their first few months of life.

Children

In preschool children, UTIs become more common and the sex ratio reverses, such that the prevalence of bacteriuria is 4.5% in girls and 0.5% in boys. In older children, the prevalence of bacteriuria falls to 1.2% among girls and 0.03% among boys. Overall, about 3–5% of girls and 1–2% of boys will experience a symptomatic UTI during childhood. However, in girls, about two-thirds of UTIs are asymptomatic. The occurrence of bacteriuria during childhood appears to correlate with a higher incidence of bacteriuria in adulthood.

Adults

When women reach adulthood, the prevalence of bacteriuria rises to between 3% and 5%. Each year, about a quarter of these bacteriuric women clear their infections spontaneously and are replaced by an equal number of newly infected women, who are often those with a history of previous infections. On average, about one in eight adult women has a symptomatic UTI each year and over half of adult women report that they have had a symptomatic UTI at some time, 20% recurrently, with the peak age incidence in their early 20s. UTI is uncommon in young healthy men, with 0.5% of adult men having bacteriuria. The rate of symptomatic UTI in men rises progressively with age, from 1% annually at age 18 to 4% at age 60.

Elderly

In the elderly of both sexes, the prevalence of bacteriuria rises dramatically, reaching 20% among women and 10% among men. In hospitals, a major predisposing cause of UTI is urinary catheterisation. With time, even with closed drainage systems and scrupulous hygiene, bacteria can be found in almost all catheters and this is a risk for the development of symptomatic infection.

Organism

E. coli causes most UTIs and although there are many serotypes of this organism, only a few of these are responsible for a disproportionate number of infections. While there are as yet no molecular markers that uniquely identify uropathogenic E. coli, some strains possess certain virulence factors that enhance their ability to cause infection, particularly infections of the upper urinary tract. Recognised factors include bacterial surface structures called P-fimbriae, which mediate adherence to glycolipid receptors on renal epithelial cells, possession of the iron-scavenging aerobactin system, and increased amounts of capsular K antigen, which mediates resistance to phagocytosis.

Host

Although many bacteria can readily grow in urine, and Louis Pasteur used urine as a bacterial culture medium in his early experiments, the high urea concentration and extremes of osmolality and pH inhibit growth. Other defence mechanisms include the flushing mechanism of bladder emptying, since small numbers of bacteria finding their way into the bladder are likely to be eliminated when the bladder is emptied. Moreover, the bladder mucosa, by virtue of a surface glycosaminoglycan, is intrinsically resistant to bacterial adherence. Presumably, in sufficient numbers, bacteria with strong adhesive properties can overcome this defence. Finally, when the bladder is infected, white blood cells are mobilised to the bladder surface to ingest and destroy invading bacteria. The role of humoral antibody-mediated immunity in defence against infection of the urinary tract remains unclear. Genetic susceptibility of individual patients to UTI has been reviewed (Lichtenberger and Hooton, 2008).

Abnormalities of the urinary tract

Any structural abnormality leading to the obstruction of urinary flow increases the likelihood of infection. Such abnormalities include congenital anomalies of the ureter or urethra, renal stones and, in men, enlargement of the prostate. Renal stones can become infected with bacteria, particularly Proteus and Klebsiella species, and thereby become a source of ‘relapsing’ infection. Vesicoureteric reflux (VUR) is a condition caused by failure of physiological valves at the junction of the ureters and the bladder which allows urine to reflux towards the kidneys when the bladder contracts. It is probable that VUR plays an important role in childhood UTIs that lead to chronic renal damage (scarring) and persistence of infection. If there is a diminished ability to empty the bladder such as that due to spinal cord injury, there is an increased risk of bacteriuria.

Babies and infants

Infections in newborn babies and infants are often overlooked or misdiagnosed because the signs may not be referable to the urinary tract. Common but non-specific presenting symptoms include failure to thrive, vomiting, fever, diarrhoea and apathy. Further, confirmation may be difficult because of problems in obtaining adequate specimens. UTI in infancy and childhood is a major risk factor for the development of renal scarring, which in turn is associated with future complications such as chronic pyelonephritis in adulthood, hypertension and renal failure. It is therefore vital to make the diagnosis early, and any child with a suspected UTI should receive urgent expert assessment.

Children

Above the age of 2, children with UTI are more likely to present with some of the classic symptoms such as frequency, dysuria and haematuria. However, some children present with acute abdominal pain and vomiting, and this may be so marked as to raise suspicions of appendicitis or other intra-abdominal pathology. Again, however, it is extremely important that the diagnosis of UTI is made promptly to pre-empt the potential long-term consequences. National guidance has been published in the UK on paediatric UTIs to promote a more consistent clinical practice by ensuring prompt, accurate diagnosis and appropriate management (NICE, 2007). A key feature of the guidance is that children with unexplained fever should have their urine tested within 24 h and attention is given to avoiding over- or underdiagnosis, appropriate investigation and the prompt start of antibiotic treatment.

Adults

In adults, the typical symptoms of lower UTI include frequency, dysuria, urgency and haematuria. Acute pyelonephritis (upper UTI) usually causes fever, rigors and loin pain in addition to lower tract symptoms. Systemic symptoms may vary from insignificant to extreme malaise. Importantly, untreated cystitis in adults rarely progresses to pyelonephritis, and bacteriuria does not seem to carry the adverse long-term consequences that it does in children.

In about 40% of women with dysuria, urgency and frequency, the urine sample contains fewer than 100,000 bacteria/mL. These patients are said to have the urethral syndrome. Some have a true bacterial infection but with relatively low counts (100–1000 bacteria/mL). Some have urethral infection with Chlamydia trachomatis, Neisseria gonorrhoeae, mycoplasmas or other ‘fastidious’ organisms, any of which might give rise to symptoms indistinguishable from those of cystitis. In others, no known cause can be found by conventional laboratory techniques. It is important to consider the possibility of urinary tract tuberculosis, as special methods are necessary for its detection. Sometimes, the symptoms are of non-infectious origin, such as menopausal oestrogen deficiency or allergy. However, most cases of urethral syndrome will respond to standard antibiotic regimens as used for treating confirmed UTI.

Elderly

Although UTI is frequent in the elderly, the great majority of cases are asymptomatic, and even when present, symptoms are not diagnostic because frequency, dysuria, hesitancy and incontinence are fairly common in elderly people without infection. Further, there may be non-specific systemic manifestations such as confusion and falls, or alternatively the infection may be the cause of deterioration in pre-existing conditions such as diabetes mellitus or congestive cardiac failure, whose clinical features might predominate. UTI is one of the most frequent causes of admission to hospital among the elderly.

Dipsticks

Dipsticks for rapid near-patient testing for urinary blood, protein, nitrites and leucocyte esterase are usually used, although there are concerns that these are reliable only when applied to fresh urine samples tested at the point of care. Assessment of colour changes on dipsticks can be subjective and automated reading systems have been developed to assist interpretation. Generally, the negative predictive value is better than the positive predictive value, so their preferred use is as screening tests to identify those specimens which are least likely to be infected and which therefore do not require culture. A perfectly valid alternative is just to hold the specimen up to the light: specimens that are visibly clear are very likely to be sterile.

The leucocyte esterase test detects enzyme released from leucocytes in urine and is ∼︀90% sensitive at detecting white blood cell counts of >10 mm⁻³. It will be positive even if the cells have been destroyed due to delays in transport to a laboratory. However, vitamin C and antibiotics in the urine such as cephalosporins, gentamicin and nitrofurantoin may interfere with the reaction. Although the presence of leucocytes is common in UTIs, it may also occur in other conditions. Particularly in children, white blood cells can be present for many reasons, including fever alone.

The nitrite test (also called the Griess test) detects urinary nitrite made by bacteria that can convert excreted dietary nitrate used as a food preservative to nitrite. Although the coliform bacteria that commonly cause UTI can be detected in this way, some organisms cannot, for example, enterococci, group B streptococci, Pseudomonas, because they do not contain the converting enzyme. In addition, the test depends on sufficient nitrate in the diet and on allowing enough time, at least 4 h, for the chemical conversion to occur in the urine. Performance of the dipstick test is generally less diagnostic in infants and younger children than in the older age groups, and this may relate in part to the small capacity and frequent emptying of the infant bladder, resulting in lower numbers of organisms and less pyuria. The use of dipsticks alone for the diagnosis of UTI is not recommended for children under 3 years of age (NICE, 2007). The inability of the test to detect group B streptococci also makes it a relatively inappropriate test for screening for asymptomatic bacteriuria in pregnancy, in which this organism assumes particular importance as a cause of neonatal sepsis.

Although a negative dipstick test for leucocytes and nitrites can quite accurately predict absence of infection, their absence does not necessarily predict non-response to antibiotic treatment and further research is needed on this (Richards et al., 2005). Some experts consider that detection of nitrites in a symptomatic patient should prompt initiation of treatment (Gopal Rao and Patel, 2009). An algorithm for the use of dipstick testing in uncomplicated UTI in adult women is set out in Fig. 36.1.

Fig. 36.1 Algorithm for diagnosis of acute uncomplicated urinary tract infection (UTI) in adult women.

There are other rapid methods for detecting bacteriuria, such as tests for interleukin-8, and no shortage of data concerning their sensitivity and specificity, but the optimal strategy will always be a compromise between accuracy, speed, convenience and cost, and is likely to be very different for different settings and populations.

Microscopy

Microscopy is the first step in the laboratory diagnosis of UTI and can be readily performed in practice. A drop of uncentrifuged urine is placed on a slide, covered with a coverslip and examined under a 40× objective. Excess white cells are usually seen in the urine of patients with symptomatic UTI, and more than 10 per high-power field is abnormal. It should be noted that there are other methods in common use, and laboratories may report white cell counts per microlitre (cubic millimetre) of urine or per millilitre. Automated machinery for microscopy of urine is increasingly used and offers increased precision and handling capacities of over 100 specimens/h. Although there is a substantial capital cost to such equipment, it is offset by savings in labour and bacterial culture materials. One feature of this equipment is that it is generally much more sensitive in detecting cells, so much so that laboratories using such systems will have a much higher number for significant results (e.g. >50 rather than >10 white cells/mm³).

It is important not to be too rigid in the interpretation of the white cell count. UTI may occur in the absence of pyuria, particularly at the extremes of age, in pregnancy and in pyelonephritis. Red blood cells may be seen, as may white cell casts, which are suggestive of pyelonephritis. As a rule of thumb, the presence of at least one bacterium per field correlates with 100,000 bacteria/mL.

Culture

Bladder urine is normally sterile, but when passed via the urethra, it is inevitable that some contamination with the urethral bacterial flora will occur. This is why it is important that laboratories quantify the number of bacteria in urine specimens. In work carried out over 40 years ago, it was demonstrated that patients with UTI usually have at least 100,000 bacteria/mL, while in patients without infection, the count is usually below 1000 bacteria/mL. Between these figures lies a grey area, and it should be appreciated that the MSU is not an infallible guide to the presence or absence of urinary infection. True infections may be associated with low counts, particularly when the urine is very dilute because of excessive fluid intake or where the pathogen is slow growing. While the quantitative criterion for ‘significant’ bacteriuria is generally taken as >100,000/mL, in some specific groups, it is less: for men, >1000/mL and for women with symptoms of UTI, it is >100/mL (SIGN, 2006).

Most genuine infections are caused by one single bacterial species; mixed cultures usually suggest contamination. If a patient is taking an antibiotic when a urine specimen is obtained for culture, growth of bacteria may be inhibited. The laboratory may perform a test to detect antimicrobial substances in the urine, and this may be useful information to clarify circumstances in which the culture is negative but a significant pyuria is present.

Non-specific treatments

Advising patients with UTI to drink a lot of fluids is common practice on the theoretical basis that more infected urine is removed by frequent bladder emptying. This is plausible, although not evidence based. Some clinicians recommend urinary analgesics such as potassium or sodium citrate, which alkalinise the urine, but these should be used as an adjunct to antibiotics. They should not be used in conjunction with nitrofurantoin, which is active only at acidic pH.

Antimicrobial chemotherapy

The principles of antimicrobial treatment of UTI are the same as those of the treatment of any other infection: from a group of suitable drugs chosen on the basis of efficacy, safety and cost, select the agent with the narrowest possible spectrum and administer it for the shortest possible time. In general, there is no evidence that bactericidal antibiotics are superior to bacteriostatic agents in treating UTI, except perhaps in relapsing infections. Blood levels of antibiotics appear to be unimportant in the treatment of lower UTI; what matters is the concentration in the urine. However, blood levels probably are important in treating pyelonephritis, which may progress to bacteraemia. Drugs suitable for the oral treatment of cystitis include trimethoprim, the β-lactams, particularly amoxicillin, co-amoxiclav and cefalexin, fluoroquinolones such as ciprofloxacin, norfloxacin and ofloxacin, and nitrofurantoin. Where intravenous administration is required, suitable agents include β-lactams such as amoxicillin and cefuroxime, quinolones, and aminoglycosides such as gentamicin.

In renal failure, it may be difficult to achieve adequate therapeutic concentrations of some drugs in the urine, particularly nitrofurantoin and quinolones. Further, accumulation and toxicity may complicate the use of aminoglycosides. Penicillins and cephalosporins attain satisfactory concentrations and are relatively non-toxic, and are therefore the agents of choice for treating UTI in the presence of renal failure.

Uncomplicated lower UTI

The problem with empirical treatment is that over 10% of the healthy adult female population would receive an antibiotic each year. The use of antibiotics to this extent in the population has implications for antibiotic resistance, a major focus of public health policy worldwide. This highlights the tension between maximising the benefit for individuals and minimising antibiotic resistance at a population level. Strategies have included diagnostic algorithms to predict more precisely who has a UTI, as well as issuing delayed prescriptions (Mangin, 2010).

Therapeutic decisions should be based on accurate, up-to-date antimicrobial susceptibility patterns. Data have been published from a European multicentre survey that examined the prevalence and antimicrobial susceptibility of community-acquired pathogens causing uncomplicated UTI in women (Kahlmeter, 2003). Among almost 2500 E. coli isolates, the resistance rates were 30% for amoxicillin, 15% for trimethoprim, 3.4% for co-amoxiclav, 2.3% for ciprofloxacin, 2.1% for cefadroxil and 1.2% for nitrofurantoin. These figures are lower than most routine laboratory data would suggest, but it should be remembered that the experience of diagnostic laboratories is likely to be biased by the overrepresentation of specimens from patients in whom empirical treatment has already failed. It is important to be aware of local variations in sensitivity pattern and to balance the risk of therapeutic failure against the cost of therapy.

Acute pyelonephritis

Patients with pyelonephritis may be severely ill and, if so, will require admission to hospital and initial treatment with a parenteral antibiotic. Suitable agents with good activity against E. coli and other Gram-negative bacilli include cephalosporins such as cefuroxime and ceftazidime, some penicillins such as co-amoxiclav, quinolones, and aminoglycosides such as gentamicin (Table 36.3). A first-choice agent would be parenteral cefuroxime, gentamicin or ciprofloxacin. When the patient is improving, the route of administration may be switched to oral therapy, typically using a quinolone. Conventionally, treatment is continued for 10–14 days.

Patients who are less severely ill at the outset may be treated with an oral antibiotic, and possibly with a shorter course of treatment. The safety of this approach has been demonstrated in a study of adult women with acute uncomplicated pyelonephritis (Talan et al., 2000). Among 113 patients treated with oral ciprofloxacin 500 mg twice daily for 7 days (± an initial intravenous dose), the cure rate was 96%.

In hospital-acquired pyelonephritis, there is a risk that the infecting organism may be resistant to the usual first-line drugs. In such cases, it may be advisable to start a broad-spectrum agent such as ceftazidime, ciprofloxacin or meropenem.

Relapsing UTI

The main causes of persistent relapsing UTI are renal infection, structural abnormalities of the urinary tract and, in men, chronic prostatitis. Patients who fail on a 7–10-day course should be given a 2-week course, and if that fails, a 6-week course can be considered. Structural abnormalities may need surgical correction before cure can be maintained. It is essential that prolonged courses (i.e. more than 4 weeks) are managed under bacteriological control, for example, with monthly cultures. In men with prostate gland infection, it is appropriate to select antibiotics with good tissue penetration such as trimethoprim and the fluoroquinolones.

Catheter-associated infections

In most large hospitals, 10–15% of patients have an indwelling urinary catheter. Even with the very best catheter care, most will have infected urine after 10–14 days of catheterisation, although most of these infections will be asymptomatic. Antibiotic treatment will often appear to eradicate the infecting organism, but as long as the catheter remains in place, the organism, or another more resistant one, will quickly return. The principles of antibiotic therapy for catheter-associated UTI are therefore as follows:

Although it often prompts investigation, cloudy or strong-smelling urine is not per se an indication for antimicrobial therapy. In these situations, saline or antiseptic bladder wash-outs are often performed, but there is little evidence that they make a difference. Similarly, encrusted catheters are often changed on aesthetic grounds, but it is not known whether this reduces the likelihood of future symptoms.

Following catheter removal, bacteriuria may resolve spontaneously but more often it persists (typically for over 2 weeks in over half of patients) and may become symptomatic, though usually it will respond well to short-course treatment.

Bacteriuria of pregnancy

The prevalence of asymptomatic bacteriuria of pregnancy is about 5%, and about a third of these women proceed to develop acute pyelonephritis, with its attendant consequences for the health of both mother and pregnancy. Further, there is evidence that asymptomatic bacteriuria is associated with low birth weight, prematurity, hypertension and pre-eclampsia. For these reasons, it is recommended that screening be carried out, preferably by culture of a properly taken MSU, which should be repeated if positive for confirmation (National Collaborating Centre for Women’s and Children’s Health, 2003).

Rigorous meta-analysis of published trials has shown that antibiotic treatment of bacteriuria in pregnancy is effective at clearing bacteriuria, reducing the incidence of pyelonephritis and reducing the risk of preterm delivery. The drugs of choice are amoxicillin or cefalexin or nitrofurantoin, depending on the sensitivity profile of the infecting organism. Co-amoxiclav is cautioned in pregnancy because of lack of clinical experience in pregnant women. Trimethoprim is contraindicated (particularly in the first trimester) because of its theoretical risk of causing neural tube defects through folate antagonism. Nitrofurantoin should be avoided close to the time of expected delivery because of a risk of haemolysis in the baby. Ciprofloxacin is contraindicated because it may affect the growing joints. There are insufficient data concerning short-course therapy in pregnancy, and 7 days of treatment remains the standard. Patients should be followed up for the duration of the pregnancy to confirm cure and to ensure that any reinfection is promptly addressed.

Cranberry juice

Cranberry juice (Vaccinium macrocarpon) has long been thought to be beneficial in preventing UTI, and this has been studied in a number of clinical trials. Cranberry is thought to inhibit adhesion of bacteria to urinary tract cells on the surface of the bladder. In sexually active women, a daily intake of 750 mL cranberry juice was associated with a 40% reduction in the risk of symptomatic UTI in a double-blinded 12-month trial. Many studies have been criticised for methodological flaws, and currently there is only limited evidence that cranberry juice is effective at preventing recurrent UTI (McMurdo et al., 2009). There have been no randomised controlled trials of the use of cranberry products (juice, tablets or capsules) in the treatment of established infection, or comparing it with established therapies such as antibiotics for preventing infection.

A hypothetical benefit in using cranberry instead of antibiotics for this purpose is a reduced risk of the development of antibiotic-resistant bacteria. A significant hazard is an interaction of cranberry with warfarin, with a risk of bleeding episodes, and available products are not available in standardised formulations. Further, cranberry juice is unpalatable unless sweetened with sugar and therefore carries a risk of tooth decay, although ironically it is reported to prevent dental caries by blocking adherence of plaque bacteria to teeth.

Antibiotic prophylaxis

In some patients, mainly women, reinfections are so frequent that long-term antimicrobial prophylaxis with specific antibiotics is indicated. If the reinfections are clearly related to sexual intercourse, then a single dose of an antibiotic after intercourse is appropriate. In other cases, long-term, low-dose prophylaxis may be beneficial. One dose of trimethoprim (100 mg) or nitrofurantoin (50 mg) at night will suffice. These drugs are unlikely to lead to the emergence of resistant bacteria, although breakthrough infection with strains intrinsically resistant to the chosen prophylactic antibiotic is possible.

Children

In children, recurrence of UTI is common and the complications potentially hazardous, so many clinicians recommend antimicrobial prophylaxis following documented infection. The evidence in favour of this practice is not strong (Le Saux et al., 2000), and although it has been shown to reduce the incidence of bacteriuria, there is no good-quality evidence that prophylactic antibiotics are effective in preventing further symptomatic UTIs and they have not been shown to reduce the incidence of renal scarring complications, which are the most important outcomes for the patient (Mori et al., 2009). Further, important variables remain to be clarified, such as when to begin prophylaxis, which agent to use and when to stop. Recent guidelines have abandoned the time-honoured recommendation for routine antibiotic prophylaxis following a first infection, although it may be considered when there is recurrent UTI (NICE, 2007).

Although evidence is limited for some recommendations, there are many common-sense general measures aimed at reducing the risk of recurrence of infection, particularly in girls. They include advice on regular bladder emptying, cleaning the perineal/anal area from front to back after toilet, treating constipation adequately and avoiding both bubble baths and washing the hair in the bath.