For the full versions of these articles see bmj.com CLINICAL REVIEW

Assessment and management of non-visible haematuria in primary care John D Kelly,1 Derek P Fawcett,2 Lawrence C Goldberg3

1Department of Oncology, Many clinicians are not sure what constitutes clinically What causes non-visible haematuria? Cambridge University, relevant haematuria; they are also unsure about when The presence of non-visible in the can have Addenbrooke’sHospital, Cambridge CB2 0QQ patients with haematuria should be referred for a transient or spurious cause; if it persists it may indicate 2Harold Hopkins Department of specialist assessment and whether they should be underlying pathology. Urology, Royal Berkshire Hospital, referred to a urologist, nephrologist, or both. Reading RG1 5AN In 2006 the National Institute for Health Research, Causes of transient non-visible haematuria 3Sussex Kidney Unit, Brighton and Sussex University Hospitals Health Technology Assessment (NIHR HTA) com- The causes of transient non-visible haematuria should NHS Trust, Royal Sussex County missioned a systematic review of the evidence for the be considered and excluded before further assessment Hospital, Brighton BN2 5BE investigation of microscopic haematuria, with a view (box 2). Transient non-visible haematuria is commonly Correspondence to: J D Kelly associated with urinary tract infection, and a repeat [email protected] to developing an algorithm for assessing patients in primary care.1 They concluded that, “Given the dipstick test after treatment for infection will determine Cite this as: BMJ 2009;338:a3021 paucity of evidence . . . it is not possible to derive an whether haematuria is persistent. Urinary tract infec- doi:10.1136/bmj.a3021 algorithm of the diagnostic pathway for haematuria tion can be the first presentation of important urinary that would be solely supported by existing evidence.” pathology, so recurrent infections are an indication for 56 None the less, the investigation of microscopic further investigation, regardless of haematuria. haematuria is important because serious underlying Observational studies in athletes confirm that repeated conditions are present in a proportion of patients. foot striking, such as in long distancerunning,can cause haematuria,7 and urine testing should be repeated at In the absence of definitive evidence, guidelines least three days after such activity. based on consensus agreement and expert opinion 34 would be useful and have been proposed. However, Spurious causes the terminology and definitions used have not been Menstruation can lead to urinary contamination with standardised, so the appropriate baseline assessment erythrocytes, so testing when menstruation has ended of patients is still unclear. In this review, we discuss is recommended. Discoloration of urine (by drugs or the rationale for introducing the terms “visible foods) and released from necrotic muscle haematuria” and “non-visible haematuria” (sympto- cells (rhabdomyolysis) are other considerations when matic and asymptomatic) (box 1). The figure shows haematuria is detected on dipstick testing.8 an algorithm for the assessment of patients with non- visible haematuria. Causes of persistent non-visible haematuria Persistent non-visible haematuria can have urological or nephrological causes (box 3). The most important Initial investigations for patients with symptomatic urological causes include cancer and calculus disease, non-visible haematuria and persistent asymptomatic which are seen in about 5% and 8% of patients, non-visible haematuria respectively.910 Urothelial cell carcinoma is the most Measure plasma and estimated glomerular common cancer detected. It is present in about 4% of filtration rate cases overall but is more prevalent in males and — Measure send a random sample of urine for increases with age to 10% in men over 60 with risk protein:creatinine ratio or albumin:creatinine ratio factors for disease.910 In young people (<40 years), (according to local practice). Twenty four hour urine especially young females, cancer is an uncommon collections for protein are rarely needed—24 hour urine protein or albumin excretion (in mg) can be approximated cause of asymptomatic non-visible haematuria, and a by multiplying the ratio (in mg/mmol) by 10 glomerular cause is more likely. It is unclear what proportion of patients with haematuria have nephro- Measure blood pressure logical as opposed to urological haematuria, because

BMJ | 24 JANUARY 2009 | VOLUME 338 227 CLINICAL REVIEW

many patients with negative urological investigations do not have a renal biopsy. However, the most Box 1 New terminology for haematuria and its diagnosis common causes are IgA nephropathy and thin Terminology membrane nephropathy.11 Visible haematuria—replaces macroscopic and gross haematuria How common is non-visible haematuria and should Non-visible haematuria—replaces microscopic and testing for haematuria be performed routinely? dipstick positive haematuria Non-visible haematuria is present in about 2.5% of Symptomatic non-visible haematuria—non-visible the general population, although it can be as high as haematuria plus lower urinary tract symptoms (hesitancy, 20%, depending on features of the study population, frequency, urgency, dysuria) or upper urinary tract such as age, sex, the presence of risk factors for symptoms 12-14 disease, and the definition used. Within cohorts Asymptomatic non-visible haematuria—incidental of patients with asymptomatic non-visible haema- detection of non-visible haematuria in the absence of turia detected by screening, the overall incidence of upper or lower urinary tract symptoms serious conditions such as urological malignancy is Diagnosis of haematuria 12 <1.5%, so the consensus is that population screen- Exclude transient causes, such as urinary tract infection, ing is not warranted. In contrast, a cause for non- before further assessment visible haematuria is found in about 15% of cases A urine dipstick test for blood is generally sufficient. It is selected for referral from primary care to haematuria 910 sensitive when performed on fresh voided urine with no clinics. These cases will usually have had an preservatives. A score of ≥1+ is positive; a trace amount is indication for urine testing, such as urinary tract considered negative symptoms.15 There is currently no evidence to A positive result for haemolysed red blood cells should be support opportunistic testing for haematuria without treated the same as for non-haemolysed red cells a clinical reason. Further assessment is warranted in patients with urinary tract symptoms and non-visible haematuria and a score of How should we test for non-visible haematuria in ≥1+ on a single blood dipstick test primary care? Urine dipstick In patients with asymptomatic non-visible haematuria confirm persistence of blood in at least two out of three Chemical dipsticks detect haem (intact red cells, free dipstick tests haemoglobin, or free myoglobin); they provide an instant result and are used to detect non-visible haematuria in It is not necessary to confirm the dipstick result by microscopy primary care.10 Chemical dipsticks are available from several manufacturers and are read visually or using automated systems on a semi-quantitative scale. haemolysed sample because we have no evidence that Although it is difficult to interpret studies on the efficiency the clinical relevance differs. of this test because of inherent design and reporting bias, analysis of pooled data sets indicates that it is a reasonable Urine microscopy way to detect non-visible haematuria in primary care counts have been used to define (positive likelihood ratio 5.99 (95% confidence interval microscopic haematuria, and cut-off points have 4.04 to 8.89), negative likelihood ratio 0.21 (0.17 to ≥ ≥ 1 varied (including 2 cells per high power field and 5 0.26)). The detection of trace haematuria can be 17 18 considered negative because the threshold for signifi- cells per high power field). Microscopy provides cance is probably less than three to five red blood cells per an accurate measure of red blood cells when assessed high power field.16 A positive result in a haemolysed by trained technicians or nephrologists in fresh sample should be treated the same as in a non- voided early morning midstream specimens of urine.19 20 However, time to analysis affects the integrity of red blood cells.21 In a prospective SOURCES AND SELECTION CRITERIA multicentre study, red blood cell counts dropped by 5-9% at five hours, 11-28% at 24 hours, and 29- We drew on evidence published in the systematic reviews 22 of the National Institute for Health and Clinical Excellence, 35% at 72 hours. Becauseimmediatemicroscopyis Health Technology Assessment (microscopic not feasible in primary care, the accuracy of haematuria),1 and guidelines for the early diagnosis and quantitative red blood cell microscopy is question- management of chronic .12 We searched able. In general practice, it is therefore not logical, electronic databases, including Medlineand the Cochrane and rarely necessary, to validate dipstick haematuria database, to identify recent publications and studies that by urine microscopy. were deemed relevant but outside the inclusion criteria of the systematic review. We included the evidence How can glomerular haematuria be distinguished from presented in published guidelines for the investigation of urinary tract haematuria? haematuria published by the American Urological Urine microscopy for red cell morphology and casts Association and the Scottish Intercollegiate Guidelines The detection of red cell casts is virtually pathogno- Network. monic for a glomerular pathology. Casts are fragile,

228 BMJ | 24 JANUARY 2009 | VOLUME 338 CLINICAL REVIEW

however, and can be very rare or absent even when glomerular disease is present. The test therefore has Box 2 Transient or spurious non-visible haematuria low sensitivity, particularly for routine samples sent Transient from primary or secondary care clinics to the Urinary tract infection laboratory. Exercise related The relevance of dysmorphic red cells as indica- Spurious tors of glomerular disease is unclear. Samples often Menstrual contamination contain a mixture of isomorphic (normal shaped) and dysmorphic cells, and the term dysmorphic Sexual intercourse includes cells with a variety of different shapes. In an Foods such as beetroot, blackberries, and rhubarb observational study to evaluate urinary particles, the Rhabdomyolysis presence of acanthocytes (erythrocytes with ring Drugs such as doxorubicin, chloroquine, and rifampicin form blebs) was significantly associated with glo- Chronic lead or mercury poisoning merular disease, but other shapes did not distinguish renal disease from urological disease.23 Urine micro- scopy relies on the examination of fresh urine by a skilled person and is most useful in specialist clinics Proteinuria rather than as a screening test in primary care. Proteinuria is a marker of glomerular damage, particularly as the amount of protein increases (low amounts can result from tubular disease or dysfunc- tion). The presence of proteinuria and haematuria increases the likelihood that the underlying disease is Visible haematuria Non-visible haematuria glomerular and is an indication for a referral to the Plasma creatinine estimated GFR Exclude transient causes nephrology department.24 The threshold for signifi- Exclude transient causes including urinary tract infection including urinary tract infection cant proteinuria in the presence of non-visible haematuria is 0.5 g/d and equivalent to a protein: creatinine ratio ≥50 mg/mmol or albumin:creatinine Symptomatic non-visible haematuria Asymptomatic non-visible haematuria ratio ≥30 mg/mmol (NICE guidelines). 2 of 3 dipstick tests positive

Yes No Hypertension Blood pressure Stop Renal disease is strongly associated with Plasma creatinine estimated GFR hypertension,25 anditmustbeconsideredinpatients Send urine for ACR or PCR with hypertension and non-visible haematuria. However, the prevalence of essential hypertension >40 years <40 years increases with age and hypertension is often an unrelated comorbidity.26 It is of most discriminatory use in people under 40, where the prevalence of Normal Abnormal hypertension is low, as is the risk of urological All of: Any one of: 910 • Estimated GFR >60 ml/min AND • Estimated GFR <60 ml/min malignancy. • ACR <30 or PCR <50 AND • ACR >30 or PCR >50 • Blood pressure <140/90 mm Hg • Blood pressure >140/90 mm Hg Do anticoagulants and antiplatelet agents cause haematuria? Urology assessment Nephrology assessment • Imaging and cystoscopy Visible haematuria in patients taking warfarin or aspirin is caused by underlying pathology in up to 25% of cases.27 Few detailed studies have assessed Cause established No cause established Cause established non-visible haematuria in patients taking these drugs. In a subgroup analysis of a prospective Primary care monitoring randomised trial to determine the effects of warfarin Annual assessment (while haematuria persists) of blood pressure, estimated GFR, and ACR/PCR and aspirin on the heart, non-visible haematuria was Referral or re-referral to urology if: detected in about 10% of patients taking the drugs, • Development of visible haematuria or symptomatic non-visible haematuria and an underlying cause including bladder cancer Referral to nephrology if: was detected in 10% of cases.28 Although not adjusted • Significant or increasing proteinuria (ACR >30 or PCR >50) for age and other risk factors, the incidence of non- • Estimated GFR <30 ml/min* • Deteriorating estimated GFR* (by >5 ml/min fall within 1 year, or >10 ml/min fall within 5 years) visible haematuria in anticoagulated patients is similar to non-anticoagulated patients, and bleeding *Confirmed on at least 2 readings and without an identifiable reversible cause cannot be attributed solely to these agents. Note: Direct referrals between urology and nephrology will depend on local commissioning guides What level of non-visible haematuria warrants further Decision algorithm for the investigation of non-visible haematuria and the referral criteria investigation? adopted by the British Association of Urological Surgeons and the Renal Association. GFR, The term non-visible haematuria best describes a glomerular filtration rate; PCR, protein:creatinine ratio; ACR, albumin:creatinine ratio clinical entity and avoids the uncertainties of either

BMJ | 24 JANUARY 2009 | VOLUME 338 229 CLINICAL REVIEW

three isolated dipstick tests assessed at an interval of FUTURE RESEARCH two to three weeks. Patients with asymptomatic Is targeted screening for haematuria effective in at risk disease are least likely to have an underlying populations, such as smokers and those with a history of urological condition, especially if they are under occupational exposure to chemicals, urological disease, 40 years. urinary tract infection, analgesic abuse, and pelvic irradiation? Future studies that define whether patients have symptomatic or asymptomatic non-visible haematuria may help redefine the indications for referral and Box 3 Causes of persistent non-visible haematuria appropriate assessment of patients with asymptomatic Urological causes non-visible haematuria Common Is there an appropriate threshold for semi-quantitative Benign prostatic hyperplasia dipstick analysis of non-visible haematuria in symptomatic and asymptomatic cases? Cancer (bladder, kidney, prostate, ureter) How do new preservation fluids affect the results of Calculus disease or nephrolithiasis quantitative and qualitative microscopy? Cystitis or Prostatitis or urethritis Schistosoma haematobium infection Less common microscopic or dipstick test definitions. We define Radiation cystitis symptomatic non-visible haematuria as haematuria Urethral strictures in patients with voiding lower urinary tract symp- toms—hesitancy, frequency, urgency, dysuria, and Tuberculosis loin or suprapubic pain in the absence of a transient Medullary sponge kidney cause such as urinary tract infection. A single Cyclophosphamide induced cystitis symptomatic episode should prompt further assess- Rare — ment in these patients serious pathology is more Arteriovenous malformation likely to be present and may cause intermittent Renal artery thrombosis bleeding. Asymptomatic non-visible haematuria is haematuria in the absence of these symptoms. When Polycystic kidney disease incidental haematuria is detected in asymptomatic Papillary necrosis of any cause patients it must be confirmed as persistent in two of Loin pain haematuria syndrome Nephrological causes Common ’ TIPS FOR NON-SPECIALISTS IgA nephropathy (Berger sdisease) Non-visible haematuria can be an indication of an Thin basement membrane disease underlying nephrological or urological condition, but Less common exclude transient causes (such as urinary tract infection) Acute glomerular disease: or spurious causes (such as menstruation) before  Postinfectious glomerulonephritis investigating  Rapidly progressive glomerulonephritis Baseline assessment of renal function (estimated  glomerular filtration rate), proteinuria, and blood pressure Systemic lupus are recommended before referring from primary to  Vasculitis secondary care  Goodpasture’sdisease All patients with visible or non-visible haematuria who  Henoch-Schönlein purpura syndrome have urinary tract symptoms should be referred for  Haemolytic-uraemic syndrome urological assessment Chronic primary glomerulonephritis: Patients aged ≥40 with asymptomatic non-visible  haematuria should be referred for urological assessment Focal segmental glomerulonephritis  Mesangio-capillary glomerulonephritis Patients under 40 with asymptomatic non-visible haematuria need urological assessment only if the  Membranous nephropathy estimated glomerular filtration rate is reduced (<60 ml/  Mesangial proliferative glomerulonephritis min) or proteinuria is >0.5 g/d Familial causes: Patients with no abnormal findings on urological  Polycystic kidney disease (autosomal dominant or assessment need long term observation, usually in recessive) primary care. Patients should be reassessed if they  Hereditary nephritis (Alport’ssyndrome) develop visible haematuria or urinary tract symptoms.  Nephrology assessment is recommended for falling renal Fabry’sdisease function or new or increasing proteinuria  Nail-patella syndrome

230 BMJ | 24 JANUARY 2009 | VOLUME 338 CLINICAL REVIEW

SUMMARY POINTS nephrology opinion is often not required either. A definitive diagnosis requires renal biopsy, and unless The terms visible haematuria should replace macroscopic or the management of the patient would be altered by gross haematuria, and non-visible haematuria (both such a diagnosis, it is hard to justify the associated risks. symptomatic and asymptomatic) should replace In patients under 40, only the presence of other risk microscopic haematuria or dipstick positive haematuria factors for severe glomerulonephritis should trigger a Urine testing for haematuria should be performed for clinical specialist nephrology assessment, namely: reasons only—current evidence does not support  ≥ opportunistic testing Proteinuria 0.5 g/d (albumin:creatinine ratio ≥30 mg/mmol or protein:creatinine ratio Thetestofchoice for diagnosinghaematuriaisurine dipstick ≥50 mg/mmol) analysis—scores of ≥1+ are positive  Estimated glomerular filtration rate <60 ml/min Transientorspuriouscausesofhaematurianeedtobe  Hypertension. excluded All patients aged ≥40 with haematuria should be Management after negative urological investigations and investigated for urological disease long term follow-up All patients with no identified urological cause should be If no abnormalities are found on initial assessment, monitored long term routine urological follow-up is not needed.29 Patients with haematuria need a nephrological assessment if they have or develop manifestations of kidney disease, Who should patients be referred to? namely24: Primary referral  Evidence of deteriorating renal function (fall in A urological cause is more likely in patients with estimated glomerular filtration rate of >5 ml/ either visible haematuria or symptomatic non- min within previous year or >10 ml/min any visible haematuria, whatever their age, and all time within past five years) those with symptomatic non-visible haematuria  stage 4 or 5 (estimated 910 aged 40 or more. In all these cases, initial referral glomerular filtration rate <30 ml/min) to urology is important to exclude malignancy.  Proteinuria (≥0.5 g/d). An exception is young adults who present with The development or presence of hypertension in cola coloured (rather than red or pink) urine, people over 40 is not a good discriminator because of particularly after a respiratory tract infection, who the increasing prevalence of essential hypertension are more likely to have acute glomerulonephritis with age, although it should be checked routinely. than urological disease. Persistent non-visible haematuria needs to be Patients under 40 with asymptomatic non-visible followed up, usually in primary care. Important but haematuria are more likely to have renal rather than undiagnosed urological disease may become more urological disease, particularly IgA nephropathy or 11 apparent through the development of urinary tract thin membrane disease. A urology referral is symptoms in previously asymptomatic patients or the unnecessary in such patients, and a specialist development of visible haematuria.29 If haematuria is caused by low grade chronic glomerulonephritis (such as IgA nephropathy), this may progress over time and ADDITIONAL EDUCATIONAL RESOURCES result in new or increasing proteinuria or worsening Rodgers M, Nixon J, Hempel S, Aho T, Kelly J, Neal D, et al. renal function (or both).30 31 Patients should therefore Diagnostic tests and algorithms used in the investigation have at least an annual review to check for new of haematuria: systematic reviews and economic symptoms and measure the estimated glomerular evaluation. 2006. www.hta.ac.uk/execsumm/ filtration rate, urinary protein, and blood pressure. summ1018.shtml The algorithm (figure) lists criteria for referral or National Institute for Health and Clinical Excellence. re-referral. Referral for suspected cancer. 2005. www.nice.org.uk/ = The joint working party of the British Association of Urological Surgeons guidance/index.jsp?action=byID&o 10968 and the Renal Association that developed the haematuria guidelines on Joint Specialty Committee on Renal Medicine of the Royal which this article is based consisted of the authors and John Anderson, College of Physicians of London and the Renal consultant urologist, Royal Hallamshire Hospital, Sheffield; John Feehally, Association. Chronic kidney disease in adults: UK professor of nephrology, Leicester General Hospital; and Robert MacTier, consultant nephrologist, Glasgow Royal Infirmary. guidelines for identification, management and referral. Contributors: DF had the idea for the article and helped write the review. JK 2006. www.rcplondon.ac.uk/pubs/contents/54942 and LG searched the literature and wrote the review. JK is guarantor. fe5-ef23-4dc7-aeb8-ed60354ffb23.pdf Funding: No special funding. NHS. Urology: 18 week commissioning pathway. 2008. Competing interests: None declared. Provenance and peer review: Offered and encouraged but not www.18weeks.nhs.uk/Content.aspx?path=/achieve- commissioned; externally peer reviewed. and-sustain/Specialty-focussed-areas/Urology

National Institute for Health and Clinical Excellence. 1 Rodgers MA, Hempel S, Aho T, Kelly JD, Kleijnen J, Westwood M. Chronic kidney disease guidelines. 2008. www.nice.org. Diagnostic tests used in the investigation of adult haematuria: A uk/Guidance/CG73 systematic review and economic evaluation. Health Technol Assess 2006;10:1-276.

BMJ | 24 JANUARY 2009 | VOLUME 338 231 CLINICAL REVIEW

2 National Institute for Health and Clinical Excellence. Chronic kidney 17 Cohen RA, Brown RS. Clinical practice. Microscopic . NEngl disease guidelines. 2008. www.nice.org.uk/Guidance/CG73. JMed2003;348:2330-8. 3 GrossfeldGD,LitwinMS,WolfJSJr,HricakH,ShulerCL,AgerterDC, 18 Sutton JM. Evaluation of hematuria in adults. JAMA et al. Evaluation of asymptomatic microscopic hematuria in adults: 1990;263:2475-80. — the American Urological Association best practice policy part I: 19 VenkatRamanG,PeadL,LeeHA,MaskellR.Ablindcontrolledtrialof definition, detection, prevalence, and etiology. Urology phase-contrast microscopy by two observers for evaluating the 2001;57:599-603. source of haematuria. Nephron 1986;44:304-8. 4 Scottish Intercollegiate Guidelines Network. Investigation of 20 Tsai JJ, Yeun JY, Kumar VA, Don BR. Comparison and interpretation of asymptomatic microscopic haematuria in adults. urinalysis performed by a nephrologist versus a hospital-based 1997. http://intranet.alemana.cl/lac_intraclinica/Mbe/GPC/ clinical laboratory. Am J Kid Dis 2005;46:820-9. Guidelines/Urologia/Hematuria.pdf. 5González CA, Errezola M, Izarzugaza I, López-Abente G, Escolar A, 21 Dowell AC, Britton JP. Microhaematuria in general practice: is urine Nebot M, et al. Urinary infection, renal lithiasis and bladder cancer in microscopy misleading? Br J Gen Pract 1990;40:67-8. Spain. Eur J Cancer 1991;27:498-500. 22 Kouri T, Malminiemi O, Penders J, Pelkonen V, Vuotari L, Delanghe J. 6 Kantor AF, Hartge P, Hoover RN, Narayana AS, Sullivan JW, Fraumeni Limits of preservation of samples for urine strip tests and particle JF Jr. Urinary tract infection and risk of bladder cancer. Am J Epidemiol counting. Clin Chem Lab Med 2008;46:703-13 1984;119:510-5. 23 Köhler H, Wandel E, Brunck B. Acanthocyturia—a characteristic 7 Bellinghieri G, Savica V, Santoro D. Renal alterations during exercise. J marker for glomerular bleeding. Kidney Int 1991;40:115-20. Ren Nutr 2008;18:158-64. 24 Joint Specialty Committee on Renal Medicine of the Royal College of 8 Sinert R, Kohl L, Rainone T, Scalea T. Exercise-induced Physicians and the Renal Association, and the Royal College of rhabdomyolysis. Ann Emerg Med 1994;23:1301-6. General Practitioners. Chronic kidney disease in adults: UK guidelines 9 Khadra MH, Pickard RS, Charlton M, Powell PH, Neal DE. A prospective for identification, management and referral. London: Royal College of analysis of 1,930 patients with hematuria to evaluate current Physicians, 2006. diagnostic practice. JUrol2000;163:524-7. 25 Buckalew VM Jr, Berg RL, Wang SR, Porush JG, Rauch S, Schulman G. 10 Edwards TJ, Dickinson AJ, Natale S, Gosling J, McGrath JS. A Prevalence of hypertension in 1,795 subjects with chronic renal prospective analysis of the diagnostic yield resulting from the disease: the modification of diet in renal disease study baseline attendance of 4020 patients at a protocol-driven haematuria clinic. cohort. Modification of Diet in Renal Disease Study Group. Am J Kidney BJU Int 2006;97:301-5. Dis 1996;28:811-21. 11 Tiebosch AT, Wolters J, Frederik PF, van der Wiel TW, Zeppenfeldt E, 26 Wolf-Maier K, Cooper RS, Banegas JR, Giampaoli S, Hense HW, van Breda Vriesman PJ. Epidemiology of idiopathic glomerular Joffres M, et al. Hypertension prevalence and blood pressure levels in disease: a prospective study. Kidney Int 1987;32:112-6. 6 European countries, Canada, and the United States. JAMA 12 Hiatt RA, Ordoñez JD. Dipstick urinalysis screening, asymptomatic 2003;289:2363-9. , and subsequent urological cancers in a population- 27 Van Savage JG, Fried FA. Anticoagulant associated hematuria: a based sample. Cancer Epidemiol Biomarkers Prev 1994;3:439-43. prospective study. JUrol1995;153:1594-6. 13 Woolhandler S, Pels RJ, Bor DH, Himmelstein DU, Lawrence RS. 28 Hurlen M, Eikvar L, Seljeflot I, Arnesen H. Occult bleeding in three Dipstick urinalysis screening of asymptomatic adults for urinary tract different antithrombotic regimes after myocardial infarction. A WARIS- disorders. I. Hematuria and proteinuria. JAMA 1989;262:1215-9. II subgroup analysis. Thromb Res 2006;118:433-8. 14 Mohr DN, Offord KP, Owen RA, Melton LJ 3rd. Asymptomatic microhematuria and urologic disease. A population-based study. 29 Mishriki SF, Nabi G, Cohen NP. Diagnosis of urologic malignancies in JAMA 1986;256:224-9. patients with asymptomatic dipstick hematuria: prospective study ’ 15 Friedman GD, Hiatt RA, Quesenberry CP Jr, Selby JV, Weiss NS. with 13 years follow-up. Urology 2008;71:13-6. Problems in assessing screening experience in observational studies 30 Geddes CC, Rauta V, Gronhagen-Riska C, Bartosik LP, Jardine AG, of screening efficacy: example of urinalysis screening for bladder Ibels LS, et al. A tricontinental view of IgA nephropathy. Nephrol Dial cancer. J Med Screen 1995;2:219-23. Transplant 2003;18:1541-8. 16 Freni SC, Heederik GJ, Hol C. Centrifugation techniques and reagent 31 D’Amico G. Natural history of idiopathic IgA nephropathy: role of strips in the assessment of microhaematuria. JClinPathol clinical and histological prognostic factors. AmJKidneyDis 1977;30;336-40. 2000;36:227-37.

A memorable 18 years The Queen Elizabeth 2 left Southampton for the last time on problems could be dealt with conservatively with intra- 11 November 2008, bound for Dubai, where she will be venous antibiotics. We gave thrombolytic treatment when moored on a palm shaped island which is to be built necessary but declined to carry pregnant ladies after specially for her in her new role as a floating hotel. She 26 weeks’ gestation: the last baby to be delivered on board worked hard in her life since her launch by Her Majesty on was in 1989, when such a restriction was not in place. Clydebank in 1967. She travelled almost 6 000 000 Helicopter evacuation, which everyone tends to think of as nautical miles, the equivalent of 14 trips to the moon and a viable option, isn’t so easy on the Atlantic as helicopters back and more than any other ship ever, with the longest can only fly out about 200 miles from base. I have used journey of almost 7000 miles non-stop returning from the them just 14 times, seven times for passengers and seven Falkland Islands after the conflict in 1982. for crew. I was the principal medical officer for most of my 18 My 18 years has been a great time of quintessential memorable years. Maritime medicine is a strange mix of Britishness at its best and wonderful to be part of—after all, general practice, emergency medicine, public health, and somebody had to go around the world with the passengers, occupational and travel medicine with some dentistry and, drink the champagne, and eat the caviar. The final evening very rarely, veterinary medicine thrown in. We are outside of 27 November 2008 in Dubai, singing all the songs from the comforting environment of a hospital, doing our own Last Night of the Proms before retiring, was astonishingly lab work and getting every result within 30 minutes with moving. I am grateful and proud I was part of it. I now point of care testing. move back to Queen Mary 2, having been on her maiden On the QE2 we could, if required, harvest four units of voyage in 2004, but I cherish memories of the most famous warm, turgid, whole, fully tested blood (sometimes laced ship in the world—the one and only, legendary, and with champagne) from passenger donors within an hour. elegant QE2. We had a digital x ray system, and, although we checked our own films, we could get a report back from a consultant Martin Carroll principal medical officer, Cunard within an hour electronically. Surgery wasn’t done as [email protected] emergencies presentedto us very early, so mostabdominal Cite this as: BMJ 2009;338:b123

232 BMJ | 24 JANUARY 2009 | VOLUME 338