Guidelines for screening with urinary dipsticks differ substantially – a systematic review

A frequently used component of general health checks is analysis of the urine [1, 2], which is often performed as a urinary dipstick test [3]. Patients admitted to hospital are also often routinely screened with a urinary dipstick, but the prevalence of this practice is unknown and likely varies between countries and regions. Use of urinary dipsticks may lead to detection of a wide array of serious conditions, e.g. urological cancers or glomerulonephritis. Early detection through screening could lead to improved prognosis, but it could also lead to unnecessary follow-up investigations such as kidney biopsies, cystoscopies, unnecessary antibiotic treatment, long-term follow-up of inconsequential abnormalities and psychological stress in healthy persons.

Dipsticks frequently combines testing for multiple substances, e.g. protein, glucose, blood, nitrite and leukocytes, which complicates the assessment of such testing. Screening for protein or albumin has been recommended for persons with certain risk factors [4-6] and is common in some countries, although there have been no trials on this [7]. In Japan, the general population has been systematically screened for proteinuria and haematuria with dipsticks for decades [8]. Enthusiasm for screening for asymptomatic microscopic haematuria has declined [9, 10], although not entirely [11, 12]. Screening asymptomatic non-pregnant persons for leukocytes, nitrite and glucose in the urine has fallen out of favour and it is unclear how often dipsticks are used for that purpose. However, it can be difficult to avoid as leukocytes and nitrite are frequently included in commonly used combined dipsticks.

There are no trials on screening for haemoglobin or protein in the urine [7, 10] and probably none on screening for glucose, leukocytes and nitrite. In other types of screening, trials have sometimes shown the benefits to be smaller than expected [13-16], and the harms greater [13, 14, 16]. In light of this lack of robust evidence, it is puzzling why screening with dipsticks is prevalent. One possible explanation may be that they are easy to use and are perceived as harmless. Furthermore, the idea that any early detection of disease is beneficial is widespread among clinicians and patients alike, despite evidence of over-diagnosis and other harms with several forms of screening [17].

It is the task of health authorities to provide recommendations on which interventions to use, both in sick and healthy people. Specialist societies also provide recommendations. The purpose of the present study was to find and describe existing recommendations on screening with urinary dipsticks, focusing on two types of screening: general health checks and routine screening of patients admitted to hospital.

METHODS

The search strategy was defined a priori, with the aim of limiting the workload while increasing the chance of finding the most important recommendations.

Six types of organisations were pre-specified: the main national health authority issuing guidance to health professionals and national professional societies for nephrology, urology, clinical biochemistry, general internal medicine, and general practice/family practice. Nine countries were pre-specified, based on the official language and on the likelihood of finding recommendations: Australia, Canada, Denmark, Ireland, New Zealand, Norway, Sweden, the United Kingdom and USA.

The internet was searched with Google to identify the relevant organisations. When two organisations of the same kind from one country appeared equally important, they were both included. When online collections of guidelines were found, e.g. the National Guideline Clearinghouse (USA) or Helsebiblioteket (Norway), these were searched, too. Other organisations were also included when judged to be important, e.g. international organisations or charities, without first looking at the contents of their website.

The website of each included organisation was browsed for guidelines or recommendations on the topic and searched using relevant pre-specified keywords, when possible. The search terms were: urinary dipstick, dipstick, urinalysis, urine strip test, urine screening, routine urinalysis, routine dipstick, routine testing, routine admission testing, admission testing, bladder cancer AND screening, (haematuria OR haematuria) AND screening, kidney disease AND screening, renal disease AND screening, proteinuria AND screening, glomerulonephritis, diabetes AND screening, bacteriuria AND screening, cystitis AND screening, health check, health evaluation, health examination, albumin. The terms were modified to suit the individual search engines and were translated when needed.

Longer documents that might have contained guidance were also searched, e.g. health technology assessments. Finally, all included organisations were e-mailed and asked whether they knew of relevant guidelines, also guidelines issued by other organisations. Recommendations were sought regarding screening with combined dipsticks and common individual components: haemoglobin, protein or albumin, leukocytes and nitrite and glucose. Recommendations for screening of specific risk groups, e.g. people with diabetes or pregnant women, were not specifically sought out. When guidance on population-based screening programmes was found, it was included as such recommendations have relevance for screening in health checks.

Relevant text, including the reference, was copied into an Excel sheet. Information on whether the included websites linked to guidelines from other organisations was also recorded along with an indication of whether the organisation explicitly endorsed that guideline. The data collection was done in November and December 2010, and in January 2013 the websites were revisited to check for new guidelines and updates.

The results were summarised in tables and in narrative. No statistics were used.

RESULTS

A total of 67 organisations were included (Figure 1, Table 1). In six cases, more than one type of organisation from a country was included, in one case two websites from the same organisation were included, and in four cases two countries shared a specialist society. Three international specialist organisations, three charities and one guideline-producing network were also included because they appeared to be important sources of guidance. Of these, five were in nephrology, one in urology and one was general.

Health checks

Combined dipsticks

No recommendations were found on screening with combined dipsticks.

Haemoglobin

Only one organisation, the UK National Screening Committee, gave a recommendation regarding screening with dipsticks for haemoglobin, recommending against using them (Table 2) [18]. Nephrological and urological societies from the UK had a joint statement recommending against testing for haematuria in the absence of identifiable clinical reasons, but did not explicitly mention dipsticks [19].

Other organisations mentioned the topic without giving recommendations. Two stated that the evidence behind screening for bladder cancer was insufficient to determine the balance between benefits and harms [20, 21], two urological societies discussed the course of action when asymptomatic microscopic haematuria had been identified [22, 23], and a list of policy positions from one public authority stated “No policy” under screening for bladder cancer, while at the same time noting that it is “very common in general practice and often part of a routine medical examination” [24].

Leukocytes/nitrite

No organisations explicitly mentioned screening with dipsticks for leukocytes or nitrite, but four organisations offered guidance on screening of healthy people for asymptomatic bacteriuria. All recommendations went against screening of non-pregnant asymptomatic persons [25-28].

Glucose

The only mention of screening for glucose with urinary dipsticks was in a health technology assessment report which noted that this technique was considered obsolete and would not be included in the report [29]. The UK National Screening Committee and a joint statement from three Danish specialist societies recommended that population screening for diabetes be avoided, without mentioning dipsticks, but both highlighted a need for increased detection of unrecognised diabetes [30, 31].

Protein/albumin

Two organisations unequivocally recommended avoiding screening with dipsticks for protein. One of these was the UK National Screening Committee, but the recommendation was found on the web page relating to screening for bladder cancer [18], while the page about screening for kidney disease did not mention dipsticks [32]. A 2008 guideline from the Canadian Society of Nephrology also recommended against mass screening with dipsticks for protein [33].

Other organisations touched on the subject without giving relevant recommendations. Kidney Disease: Improving Global Outcomes noted that there appears to be no evidence for screening unselected populations with reagent strips [34].

The Scottish Intercollegiate Guidelines Network noted that dipstick testing can be used to identify persons at risk of subsequent end-stage renal disease and cardiovascular disease, but also noted that “urine dipstick testing cannot be used reliably in isolation to diagnose the presence or absence of proteinuria” [35]. A New Zealand public authority gave its policy regarding screening for chronic kidney disease as “opportunistic screening and self-testing using a urinary dipstick” [36].

Several other organisations, including the influential National Kidney Foundation K/DOQI guideline, gave no recommendations for or against general screening, but recommended screening high-risk groups for chronic kidney disease, with varying definitions of what constituted high risk [37-42]. The recommended tests were typically measurement of the albumin-creatinine ratio (ACR) or an albumin-specific dipstick in combination with the estimated glomerular filtration rate. The topic of ACR dipsticks was mentioned by the National Institute of Health and Clinical Excellence [37], stating that dipsticks should only be used if they are capable of measuring albumin at low concentrations and of expressing the results as an ACR.

Admission to hospital

No recommendations were found on any kind of routine dipstick screening on admission to hospital.

DISCUSSION

Recommendations on the use of urinary dipsticks for screening purposes were scarce and often unclear. Despite a thorough search of websites from health authorities and medical societies in nine countries, no recommendations were found on the use of combined dipsticks in health checks or at admission to hospital.

Only one clear statement was found on screening for microscopic haematuria with dipsticks, recommending against their use. Surprisingly, only one urological society gave clear guidance on screening for microscopic haematuria, recommending against, but did not mention dipsticks. Other organisations discussed the topic without giving recommendations. The scarcity of clear guidance may be related to the fact that the literature seems to be in a stalemate, with some observational studies hinting at a possibly important beneficial effect [8, 11], but with no trials to confirm or refute this.

No clear recommendations were found on screening for urinary glucose with dipsticks, but, as was stated in one health technology assessment report, this technique is considered obsolete. It is likely that some experts consider it self-evident that it should not be used, but it is unlikely that all practitioners – including nurses who perform the tests in hospitals – know this.

Regarding screening for bacteriuria, only four recommendations were found, and they all clearly discouraged this practice, except in pregnant women. However, none of the recommendations specifically mentioned dipsticks as the screening method.

Screening for chronic kidney disease was frequently mentioned, and some organisations discussed limitations of dipstick testing for protein, but clear recommendations were scant. As with glucose, it is possible that some experts simply consider dipstick screening for proteinuria an obsolete technique not worth recommending against in guidelines. Assessing the albumin-creatinine ratio in high-risk persons was often recommended, but although this is a better measure than proteinuria, and although a high-risk only strategy likely reduces over-diagnosis and overtreatment, it is still not clear whether screening is beneficial or not. Albumin-creatinine ratio and dipstick proteinuria are predictors for total and cardiovascular mortality [43], but ACR only adds minimally to traditional cardiovascular risk prediction methods [44]. Treatment with angiotensin-converting enzyme inhibitors appears to reduce end-stage renal disease in persons with chronic kidney disease, macroalbuminuria and diabetes [7], but has not been proven effective for non-diabetic chronic kidney disease stage 1-3, which constitute the majority of cases [45]. Screening trials have not been conducted and information on the harms of diagnosis, treatment, and follow-up is scarce [7].

The comprehensive and systematic search used in this study far exceeds what can be expected from a clinician looking for guidance. However, it is possible that some guidance has been overlooked or misinterpreted. The language limitations and the selection of certain medical fields probably reduced the number of recommendations found. Also, the choice of not searching regional and local authorities may mean that some guidance has been missed. However, such guidance, if it exists, will not necessarily reflect any national or international consensus. Four hospitals were contacted and none of them had any policy on the topic.

The combined dipsticks in common use in health checks and at admission to hospital have a potential to do harm, as do all medical interventions. Even when used for non-screening purposes, they give redundant information that may initiate a diagnostic cascade, and from this viewpoint their existence can be questioned. Using them for screening purposes without solid knowledge from randomised trials that the benefits exceed the harms is unethical, and guidance from authorities and specialist societies should reflect this. There is a need for clear and pragmatic “Do not use” lists regarding tests, helping practitioners avoid subjecting their patients to possibly useless and potentially harmful tests.

Correspondence: Lasse T. Krogsbøll, The Nordic Cochrane Centre, Rigshospitalet, 2100 Copenhagen Ø, Denmark. E-mail: ltk@cochrane.dk.

Accepted: 3 December 2013

Conflicts of interest:Disclosure forms provided by the authors are available with the full text of this article at www.danmedj.dk

Acknowledgements: The project arose from discussions with Peter C. Gøtzsche and Karsten Juhl Jørgensen, who also commented on the protocol and the manuscript.