Diversity and Complexity of Urinary Tract Infection in Diabetes Mellitus

Lukman M. Hakeem, Diptendu N. Bhattacharyya, Cyril Lafong, Khalid S. Janjua, Jonathan T. Serhan and Ian W. Campbell


British Journal of Diabetes and Vascular Disease. 2009;9(3):119-125. 

In This Article


In most prospective studies patients with diabetes are characterised as having 'complicated infections'. However, categorising all patients with diabetes into this group trivialises the concept.

A 'complicated' UTI diagnosis should be reserved for diabetic patients who also have metabolic/systemic disease such as ketoacidosis, hyperosmolar state, renal impairment or structural abnormalities such as neurogenic bladder, renal perinephric abscess, papillary necrosis, renal/bladder emphysema, renal calculi, obstruction and urethral catheterisation. UTI in pregnancy and patients who relapse after therapy due to microbial persistence in renal or prostatic parenchyma also fall into this category.[4,44] Those diabetic subjects with more remote or metastatic infections such as pneumonia (see figure 1) are also in this category.

Initial management consists of hydration usually with intravenous fluids for which the patient is hospitalised, and intravenous antibiotics, in addition to aggressive control of hyperglycaemia. Antibiotics that could be used for UTIs and complications are shown in table 1. Treatment protocols should avoid nephrotoxic antimicrobial agents whenever possible. Otherwise treatment regimens selected for patients with acute cystitis and pyelonephritis in patients without diabetes are appropriate for patients with diabetes. However, most authors prefer antimicrobial agents which achieve high levels not only in the urine but also in the renal tract tissues, such as fluoroquinolones, trimethoprim-sulphamethoxazole (TMPSMX) and amoxycillin-clavulanic acid.[25,45] This may hold especially true given the data indicating invasion of E.coli into the bladder cells.[46] In this country, however, trimethoprim is commonly used as a single agent treatment for UTIs instead of TMP-SMX. Few therapeutic trials have specifically been performed with diabetic patients. Due to frequent upper urinary tract involvement and possible serious complications many experts recommend a 7–14-day oral antimicrobial regimen for bacterial cystitis in diabetic patients, instead of the usually recommended 3-day course.[25,45,47] The standard duration of therapy for uncomplicated pyelonephritis in both diabetic and non-diabetic patients is 14 days.[30,45,48] However, studies have shown a 7-day course of oral ciprofloxacin to be adequate and effective for uncomplicated pyelonephritis.[49] Vigilance for complications must occur throughout the care of an acutely ill patient with UTI. As these complications are common in patients with diabetes, their anticipation can lead to earlier interventions and fewer serious adverse outcomes.

There are increased concerns about emergence of resistant organisms in patients being treated for recurrent UTIs. Since UTIs are common in diabetes this becomes a major concern in this group of patients. Prominent among such infections are organisms producing ESBLs which have the ability to hydrolyse oxyimino-cephalosporins and monobactams. Enterobacteriaceae, especially Klebsiella spp. producing ESBLs such as SHV and TEM type enzymes, have been established since the 1980s as a major cause of hospital-acquired infections. However, more recently, enterobacteriaceae (mostly E. coli) producing novel ESBLs, the CTX-M enzymes, have been identified predominantly from the community as a cause of UTIs.[50] Resistance to other classes of antibiotics, especially the fluoroquinolones and aminoglycosides, is often associated with ESBL-producing organisms. Therefore the choice of antibiotics that could be used becomes severely restricted and may lead to increased prescription of more broad-spectrum and expensive drugs such as the carbapenems.[51] A heightened awareness of these organisms by clinicians and enhanced testing by laboratories, including molecular surveillance studies, is required to reduce treatment failures, to limit their introduction into hospitals and to prevent the spread of these emerging pathogens within the community.

Figure 3.

Repeat CT scan of abdomen with contrast showing marked improvement of the perinephric abscess with reduction in size of the collection 2 weeks after percutaneous drainage. Abscess fluid showed an isolation of Enterococcus faecalis, group G streptococci and Bacteroides species. The patient continued with the antibiotic regime

Upper UTIs and disseminated infections due to fungi require systemic therapy. The appropriate treatment of candida infection confined to the bladder remains controversial.[40] Distinguishing such infections from colonisation is often difficult. The presence of symptoms or pyuria suggests infection. Spontaneous resolution of funguria occurs in many cases.[52,53] Removal of an indwelling catheter, if one is present, is recommended as the initial intervention. The treatment options include bladder irrigation with amphotericin B,[54] a single dose of intravenous amphotericin B,[55] or oral fluconazole.[56] Currently fluconazole is the preferred drug of choice because of its ease of administration and relative absence of toxicity.

Surgical management may be necessary in patients with renal emphysema and suppurating renal or perinephric abscesses. Obstruction should be sought and treated appropriately. The traditional treatment of emphysematous pyelonephritis is nephrectomy of the affected kidney. Surgery has been reported to lower the mortality from 80% in patients treated with antimicrobial treatment alone, to 20%.[47] Increasing numbers of cases are reported of successful antibiotic therapy combined with radiographically guided percutaneous drainage in cases where infection is localised.[57,58] Therefore initial conservative management with CT-guided drainage should usually be attempted with sequential studies to ensure that infection is controlled, air, if present, is evacuated and the patient is adequately responding.[4] Figures 2–4 illustrate a case for which a perinephric abscess was successfully treated with percutaneous drainage. Total nephrectomy is considered for patients whose condition does not improve clinically or in whom gas spreads despite non-surgical treatment.

Figure 4.

Repeat CT scan 3 months later showing marked improvement with significantly smaller residual collection around the right kidney which was thought to represent a haematoma. The patient completed a total 4-week course of antibiotics and made a good recovery

Many UTIs in patients with recurrent bacteriuria can be prevented through strategies which include complete emptying of the bladder during voiding, antimicrobial prophylaxis, less use of spermicides, and optimal catheter care.[59] These are important points to consider in patients with diabetes, and this advice should be provided to all patients during diabetic education programmes and particularly after an initial episode of UTI.

In conclusion, patients with diabetes have a higher incidence of both asymptomatic bacteriuria and symptomatic UTIs, which often lead to complications both local and remote as illustrated in the two case histories presented. Management regimens for the most part are not evidence based. Despite the clinical and economic significance research interest on this subject has been inadequate and therapeutic trials enrolling patients with diabetes are necessary to define the best therapeutic options, antimicrobial agents and optimal treatment duration. Clinicians should respond proactively to give optimal treatment to reduce morbidity and mortality in diabetic patients with UTIs.