Clinical review

Asymptomatic haematuria ... in the doctor

Chris Del Mar, professor. 

Centre for General Practice, University of Queensland, The University General Practice, Inala, Queensland 4077, Australia

c.delmar{at}mailbox.uq.edu.au

The patient was waiting in the consulting room; everything was nearly ready. The occasion was the examination in general practice for fifth year medical students. We run an objective structured clinical examination. For this part, the student had to measure the patient's blood pressure (the "patient" was actually someone recruited from our general practice), test his urine using a dipstick, and report to the examiner within the five minutes between bells. Just one thing was missingthe midstream sample of normal urine for testing. Because I did not want to disturb the volunteer patient, I collected it from myself. I measured the patient's blood pressure again (this had to be done after every 10th student)it was stable. And I tested the urine to check it was normalit was not.

For the next two hours, students either told me (or, in the case of those less skilled at this technique, did not tell me) that there was a trace of blood in the urine. This was not a problem as far as the examination was concerned because the marking was not affected by the test result. But it was a problem for me. What should I do? I tested my urine again a week later, and when I found it was still positive I sent a specimen to the laboratory. The report stated that urine culture was negative but confirmed the presence of normal red cells (30/ml).

Conventional medical teaching had taught me that bleeding must come from somewhere. The model that sprang to mind first is summarised in the table. I then checked with a textbook of surgery.¹ I had forgotten tuberculosis and schistosomiasis as causes of haematuria. A textbook of medicine² suggested further assessments, including checking my blood relatives for urine abnormality and carrying out haemoglobin electrophoresis and 24 hour urinary estimations of urate and calcium excretion. If all these investigations were negative, intravenous urography, cystoscopy, and renal computed tomography were proposed, with indefinite regular follow up thereafter. The essential feature of this model is that identifying the lesion anatomically or physiologically is the key to managing the problem. Early diagnosis of some of the serious causes of haematuria such as transitional cell carcinoma might affect the outcome favourably by enabling treatment to be given. With other causes such as minimal change glomerulonephritis, early diagnosis is unlikely to affect mortality or morbidity.

I consulted my general practice colleague. His approach was similar. He ensured that I had checked my urine microscopy and culture to establish whether the blood was haemolysed or not, and he measured my blood pressure. He wondered whether my bicycle riding might be the cause, and suggested I recheck the urine after a month or two. If test results were still positive, it looked as if the cascade of likely events would include ultrasonography, urine samples for malignant cells, and then probably referral to a urologist for consideration of cystoscopy and intravenous urography.

"We probably won't find any cause for it," he said. I think this was to reassure me. I had time to think about adopting an evidence based approach.

	Formulating the question

Top Formulating the question Searching for evidence The evidence The clinical decision Discussion References

The most difficult part of adopting evidence in practice is formulating the question. This involves quite a different way in thinkingaway from "patho-anatomical-physiological" questions towards empirical ones. What is the chance of having a serious condition with asymptomatic haematuria? What sort of study did I want? Ideally, it would be huge study of general practice academics in their 40s who had unexpectedly found haematuria during their academic duties. And the outcome would be a cohort study with a long follow up to see what illness occurred.

	Searching for evidence

Top Formulating the question Searching for evidence The evidence The clinical decision Discussion References

The standard textbooks on my shelf were no help in providing the evidence, so I consulted The Cochrane Library (on CD Rom). This would have been the most convenient source of data and is strong on evidence for different treatments, but unfortunately it does not yet include routinely collected data on the course of diseases and conditions. Since The Cochrane Library did not help, I decided to search Medline. My search strategy was simple,³ and probably sensitive at the expense of being less specific: the term used was "(incidence or explode (mortality or (follow-up studies) or mortality or prognosis: or predict: or course)) and (hematuria or haematuria)."

Inspection of titles helped me discard immediately about half the 230 hits, but reading through the printout of abstracts of the remainder took an evening. There were no systematic reviewsthe best form of evidence.⁴ Two articles were clearly useful because they described large studies with long term follow up of people with negative and positive dipstick test results. I obtained the full versions of the papers from the library.

	The evidence

Top Formulating the question Searching for evidence The evidence The clinical decision Discussion References

In a British study, 2.5% of more than 10 000 men screened for asymptomatic haematuria by dipstick testing had positive results, of whom 60% were investigated further by their general practitioners.⁵ Three had a serious condition that was amenable to curetwo had bladder cancer and one had reflux nephropathy. This study seemed to be fairly close to my ideal. It gave a prognosis of the outcome someone like me (my situation being similar to screening) would expect. It suggested that I was unlikely to have a serious condition that was amenable to cure. Of course, even this may be an overestimate of the benefits of screening. Perhaps those three people would have developed symptoms such as frank haematuria or dysuria sufficiently early to negate the beneficial effect of screening on their prognosis.

Another study was done in California.⁶ Over 20 000 middle aged people were screened by dipstick for haematuria. An unexpected positive result was found in nearly 3%, 99% of whom were followed up. Over the next three years, three patients developed urological cancers (two prostatic and one bladder). This study is more relevant because it looked at the outcome of people whose dipstick test was not positive; their probability of developing urological cancer was no less than that of people whose dipstick test was positive. According to this study, the likelihood of my developing urological cancer was 0.5%, whether I had haematuria or not.

Were the studies of good enough quality? As they were studies of patients selected for screening, they were probably the best match with my situation that I could hope for. The second study, particularly, seemed to address a variety of potential selection biases. In the absence of a formal meta-analysis, this seemed to be a sufficiently serious "dip" into the published reports to answer my clinical question.

Another study on young men in the air force shed a little light on factors associated with microscopic haematuria. Neither exercise, recent sexual intercourse, nor flying were associated with microscopic haematuria, although recall of a history of urethritis was.⁷ My bicycling was probably unrelated.

	The clinical decision

Top Formulating the question Searching for evidence The evidence The clinical decision Discussion References

What should I do now? I decided that the chance of having an adverse outcome was not sufficiently high for me to bother with further investigations. I would adopt a management policy of "expectant observation." I have watched for frank haematuria or any other new signs, and I check my urine every three months (the haematuria has remained, but there has been no increase in the concentration of red cells over the past year).

I have applied my own values to the clinical decision. If, in similar circumstances, a patient of mine elected to proceed with further investigations, I would comply. This does not address matters of gatekeeping (resource allocation), which probably should be dealt with away from the consultation.

	Discussion

Top Formulating the question Searching for evidence The evidence The clinical decision Discussion References

How has an evidence based approach helped? The main difference was the change in clinical thinking that allowed me to break away from the patho-anatomical-physiological approach and adopt an empirical one. These steps are not easy. Searching the published reports is still awkward and time consuming. Some answers are difficult to find. How long, for example, should we carry on looking before concluding that there seems to be no published work to guide us? We also need a forum of peers and those skilled at evidence based medicine in which test out our ideas so that we can reassure ourselves that we are not completely off course. If health authorities are serious about promoting evidence based medicine in clinical practice, they may have to consider providing a service (perhaps like pathology, radiology, or referred specialist opinions) to help clinicians to take these steps.

	Acknowledgments

Paul Glasziou constructively read earlier drafts and checked, with Geoff Hirst, that I was not completely off course; my thanks to both.

	Footnotes

Funding: None.

Competing interests: None declared.

	References

Top Formulating the question Searching for evidence The evidence The clinical decision Discussion References

1.	Tipcraft RC. Urinary symptoms. Investigation of the urinary tract. Anuria. In: Mann CV, Russell RCG, William NS, eds. Bailey and Love's short practice of surgery. London: Chapman and Hall Medical, 1995:904-914.
2.	Denker BM, Brenner BM. Cardinal management of renal disease. In: Fauci AS, Braunwald E, Isselbacker KJ, Wilson JD, Martin JB, et aleds. Harrison's principles of internal medicine. New York: McGraw Hill, 1998:258-262.
3.	Del Mar C, Jewell D. Tracking down the evidence. In: Silagy C, Haines A, eds. Evidence based practice in primary care. London: BMJ Publishing, 1998:21-33.
4.	Sackett DL, Richardson WS, Rosenberg W, Haynes RB. Evidence-based medicine. London: Pearson Professional Publishing, 1997.
5.	Ritchie CD, Bevan EA, Collier SJ. Importance of occult haematuria found at screening. BMJ 1986; 292: 681-683.
6.	Hiatt RA, Ordonez JD. Dipstick urinalysis screening, asymptomatic microhematuria, and subsequent urological cancers in a population-based sample. Cancer Epidemiol Biomarkers Prev 1994; 3: 439-443[Abstract].
7.	Froom P, Gross M, Froom J, Caine Y, Margaliot S, Benbassat J. Factors associated with microhematuria in asymptomatic young men. Clin Chem 1986; 32: 2013-2015[Abstract/Free Full Text].

(Accepted 19 April 1999)