The main finding of this study is that agents frequently prescribed to prevent UTIs in Norwegian nursing homes lack documented efficacy in the elderly. Recommended agents like trimethoprim, nitrofurantoin and vaginal estrogens are infrequently used. Instead we observed a high prevalence of methenamine, vitamin C, systemic estrogens and cranberry products. We conclude that prescribing of prophylactic agents for UTIs in nursing homes is not evidence-based according to the literature and current national guidelines.
Recurrent UTIs are common, especially in older women. Thus, high frequency of residents using prophylaxis in our study could be expected. However, the high variation in prevalence and the choices of prophylactic agents were surprising. One or several concomitant prophylactic agents were used by almost one fifth of the residents. At present we do not know which factors contribute to this high variation in prophylactic prevalence.
This and former studies have shown that the urinary antiseptic agent methenamine is frequently used in Norway and in Norwegian nursing homes, in contrast to most other countries in Europe [18
]. In Norway, the use has even been increasing, and in 2010 methenamine represented 17% of antibacterials for systemic use, measured as share of DDDs among antiinfectives for systemic use (ATC-group J01) [20
]. A Cochrane review of methenamine for preventing UTIs, found the overall quality of the studies to be poor. Few studies addressed long term use or the use in postmenopausal women or elderly in general. In the review it was concluded that methenamine may be effective for preventing UTI in patients without renal tract abnormalities, particularly when used for short-term prophylaxis [15
]. In our study we found methenamine to be used frequently. The use appeared to be continuous which is in accordance with the SPC that does not state any limitations to treatment duration but in contrast with other documentation as summarized in the Cochrane review [15
]. Current Norwegian guidelines include prophylactic use of methenamine as in patients without catheter, but point out the low grade of documentation for this agent [3
Traditionally, cranberries have been used to prevent UTIs, but studies of efficacy have shown conflicting results. A recent Cochrane review concluded that cranberries could be effective, but that the evidence for the elderly still was inconclusive [16
]. Another review did not recommend cranberry products for the prophylaxis of UTIs due to heterogeneity in study design and results, and a lack of consensus regarding both dosage regime and formulations. Interactions may be a problem in patients with polypharmacy, especially for concentrated cranberry products [22
]. In addition, intolerance to cranberries probably represents a problem in the elderly and high withdrawal rates are reported in several studies [16
]. In contrast to this, the withdrawal rates due to adverse reactions were the same comparing 500 mg cranberry extract with 100 mg trimethoprim in one study and with trimethoprim-sulfamethoxazole in another study. These studies included community dwelling women, age 45-93 years and 18 years to menopause respectively, and the relevance for nursing home residents is unclear [23
Vaginal estrogens have been shown to decrease UTIs while systemic estrogens do not appear to have the same effect [13
]. In our study only 31% (25 of 81) of the estrogens prescribed as UTI prophylaxis were for vaginal administration. This is troublesome because systemic estrogens have been associated with increased risk of cardiovascular disease, venous thromboembolic events and breast cancer [25
]. Current Norwegian guidelines recommend vaginal estrogens to women with recurrent UTIs [3
Trimethoprim and nitrofurantoin were the two antibiotics prescribed for prevention and used only by few of the nursing home patients. Long-term antibiotics are well documented to reduce the rate of UTIs but may be complicated by bacterial resistance and adverse drug reactions (ADRs) [11
]. The prevalence of trimethoprim resistance in Escherichia coli
isolates in Norway was 19% in 2007, being the drug with the highest prevalence of resistance in urinary tract isolates [9
]. Nitrofurantoin was only resistant in 2.3% of the isolates [9
]. However, nitrofurantoin should be used with caution in patients with renal impairment as reduced renal clearance increase the risk of ADRs and sufficient concentration in the urine depends on renal function. In addition, long term use of nitrofurantoin is associated with lung fibrosis and peripheral neuropathy [3
]. As glomerular filtration rate decline by age, nursing home patients are at risk of a negative risk/benefit balance for the use of nitrofurantoin [27
]. Local resistance pattern and individual renal function should therefore be considered before prescribing trimethoprim or nitrofurantoin for long-term use to nursing home residents.
Vitamin C has traditionally been regarded as effective in preventing recurrent UTIs. However, we could not find any studies showing that vitamin C is effective in preventing UTIs in the elderly. Norwegian guidelines do not recommend the use of vitamin C for UTI prophylaxis in nursing homes.
To summarize, our results show lack of evidence-based prescribing of prophylactic agents for UTIs in Norwegian nursing homes, according to evidence in the literature and current national guidelines. Low dose antibiotics and vaginal estrogens are at present recommended agents. However, they were only used by 18% of the residents prescribed prophylaxis in this study. The use of methenamine and cranberries were common, but the efficacy of these agents is not well documented. Systemic estrogens and vitamin C have no place in preventing UTIs and are to be considered as inappropriate as alternative and more evidence-based therapy exists. Prophylactic agents were prescribed continuously to residents, suggesting that their use were not regularly evaluated. The influence of catheter use and other risk factors are not known in our study, but an European surveillance study found the use of urinary catheter to be low (4,8%) in Norwegian nursing homes [28
Generalisations from prevalence studies must be made carefully, but results can be useful to define quality improvement projects in institutions with less developed infection control systems than in hospitals. In previously published data from this study we showed prescribing of antibiotics for therapy to be in line with other studies from nursing homes [6
]. We found all prescribing of UTI prophylaxis to be continuously which minimize the variation in results of this prevalence study. Interestingly, a recent Dutch study found a high occurrence of non-catheter related UTIs in nursing homes, perhaps due to frequent faecal incontinence among residents [29
]. If this is true, prophylaxis should be focused on improving hygiene and providing incontinence materials rather than prescription of the agents found in our study. Notably, our study included a large number of nursing homes representing diversity in size and function. Thus, we believe the results from this study apply to Norwegian nursing homes in general, and that they also could be of international interest.