36. Urinalysis paralysis

I'm doing fine .

So , yeah , it came back to driving stick and yeah , it's just got you a bit down , I guess , having to go back to driving stick . You're in a dip the stick , you're in dip stick Anyway .

So , Callum , tell me what . How would you diagnose a UTI in a patient ?

But if you look at various populations , cohort studies where they took patients and asked them their symptoms and then followed that through to a urine culture and got a gold standard diagnosis that the main symptoms that were predictive of people having a UTI or negatively predictive so they had a lower likelihood ratio If they didn't have them was dysuria , urinary

frequency so having those would increase your likelihood of living at a UTI and also urinary urgency and things like vaginal discharge were negatively associated with having UTI . So if you had those symptoms you're less likely to have a UTI . So that's from Geisen et al .

And actually if you look at the combination Geisen et al , there we go we can't even agree on the name for our investigations . And another study bent it out in 2002 , they looked at patients . If you had a combination of dysuria and urinary frequency and no vaginal discharge , this is moving . Again , you had a positive likelihood ratio of 24.6 .

So essentially what that means is you're a large increase in the likelihood of you having a UTI with those symptoms . You work out a likelihood ratio . A positive one is sensitivity over one minus specificity . So it's a quick way of combining those two factors . So really that's huge .

So the end of the most patients with a UTI , and it's straightforward they come in , they've got dysuria , they've got urinary frequency . It doesn't look like something else , it looks like UTI , smells like a UTI maybe , and it tastes like you . Now I'm going to say that .

So we don't taste urine anymore and you give them antibiotic space in your guidelines , or , if it's more complicated , then you might take some more tests .

If you don't have it , you're more likely to have it than if you did have it . Exactly , and this is what the Bentatal study did they mixed dysuria and frequency with absence of vaginal discharge and got a likelihood ratio of 24 , which is very impressive , and then the reverse of that had a likelihood ratio of 0.3 .

So if a likelihood ratio is less than one , it's the presence of that constellation of symptoms or diagnostic test is making it less likely that you've got the disease . So that's all to the good . Those are the symptoms that we look for of it .

But I imagine that quite a lot of people , if you ask that question , would have said I'll do your analysis and check for what sells and nitrites . So I think it's first worth going in now and talking about dipsticks .

And in Hippocrates it always pops up 400 BC , looking at color . Your order changes . And throughout history since then there's been various advancements through hundreds of years where people have looked for different things in urine and associated them with different diseases .

And the main reason we're mentioning history is not because it's particularly relevant to your learning , but just because there was some , let's just say , quacks Is that the right term , I'm not sure People that were trying to take advantage of this fascination of urine , and there was a Euroscopist in the Middle Ages who decided that they could diagnose all sorts of

things , from health to disease , to the future , from urine , and so they were labeled as piss prophets , which I think is just such a great name . I love that . I love that .

Sometimes they have different tests the ones that we use in the UK , the Siemens ones , and they've got other things like protein and blood which can be useful for picking up chlamydia , oenophritis , and we've also got ketones , used for diabetes , sugar , glucose , bilirubin , specific gravity , which everyone looks at and wonders what it is , and pH , which can be useful ,

but for the purposes of infection . We're thinking about leukocyte esterase , or leukocytes or white cell count , as I sometimes refer to , and nitrites , but nitrates . And essentially , if you're not familiar with this test , you it's very easy .

You get a clean catch urine or some urine sample that you've collected and you put it into a little bottle , you put your dipstick in and then you wait the allotted amount of time . So in this one , two minutes for leukocytes and 60 seconds for nitrites , and then it gives you a color change and you read that against a scale .

So it's not , it's not , it's like a semi quantitative . You know you're not got a natural measurement .

But what we are interested in is the top two sort of things , which are nitrites and leukocytes . Why don't you take us through what they are looking for next ?

And then you wonder , you know , ask why ? Oh , they smell urine . So I did the urinalysis , which is always a bit like okay , why did you do your urinalysis ? And it was positive , so they've got UTI . And then you're like okay , what did the urinalysis show ? Oh , garden protein . And you're like what about leukocytes and nitrites ? No , Well , okay .

Because not only are we about to absolutely reign on the parade of urinalysis in general , but we're about to say that leukocytes and nitrites don't have a role in a bunch of situations , and sometimes they're not the things that are positive .

Sometimes there's other stuff that's basically got nothing to do with the presence or absence of infection , which is really what we are focusing on here .

And that indicates back to urea or urinary tract infection or some other cause of inflammation in the urinary tract . So it's important to note that this is a sensitive test and it's quite specific . But there are other reasons to have leukocyte esterase in the urine .

And that's what the high sensitivity of troponin is . It's sensitive at detecting MI . Specificity is the opposite of that In your population , people that don't have the disease , how many of them would test negative ? And so the troponin here can be used as an example too , because although it is very sensitive , it's not very specific .

It goes up in lots of things . If you run a marathon , it will go up . If you have myocarditis it will go up . If you have enocarditis it will go up . If you have sepsis it will go up . So there's lots of people in the don't have an MI population that will have a raised troponin for other reasons , and so the specificities are about 70 to 80% .

And so that's sensitivity is disease positive that tests positive , and specificity is disease negative , that would also test negative .

So if you find nitrites , there's not much else that's going to cause that other in bacteria .

And so if we come on to this , we've got a reference for this as well , but it's quite an old paper from 1999 , but they basically looked at 400 , just under 500 patients , women that had a suspected uncomplicated UTI , and then they got their urine cultures and classified them by having , like , how strongly positive the urine culture was , so ranging from 10 to

the free calling for our units per mil , to 10 to the five . And they looked at the sensitivity , specificity of leukocyte estuaries , nitrites and then the two of them combined and also the positive predictive value and the negative predictive value of these two tests .

And essentially , as you might expect , leukocyte estuaries was more sensitive and was picking up , was positive essentially , even with the tend of the freeze , the lower levels of bacteria , whereas nitrites was often not picking that up . So it was more sensitive .

And if you combine the two of them so say you took leukocyte estuaries and nitrites and you had colony counts of 10 to the 5 , so a good going , proper UTI . It was 84% sensitive and the two of them combined , specificity was 98.3% .

So combining the two , you get the strengths of both , but like none of the deficiencies .

So you were missing more by having that bar set to help both of them .

To put that into words in terms of the data , the nice guidelines are quite clear . For women that have mild symptoms , not catheterized , we'll come to that in a second dipstick the urine and they say if both dipstick tests are negative , uti unlikely , makes sense . If leukocyte estuaries is positive alone , uti is moderately likely .

If the nitrite test is positive , with or without positive leukocyte estuaries , a UTI is highly likely because it's so specific . I don't think you often see that .

I think you usually see the leukocyte estuaries being positive , but putting more weight on nitrite as that sort of rule in test and for women who have a moderate or severe symptoms , with or without a catheter , make a working diagnosis of UTI without doing a deuronosis .

So you're only really using the urinosis for women with mild symptoms of UTI and in men it just says don't use urine dipstick test .

So , in terms of your clinical reasoning , as a methodology for examin' how we think as clinicians , which is really interesting when you I think it's easy to on paper say objectively , like this , what influenced my decision making , just like the free drug rep lunch won't affect my prescribing .

So , you know you can look on paper and say like , actually you know we'll come onto this in a little bit about like positive and negative predictive values of tests and say , ok , well , I know that it's 50-50 , so it's actually influenced my decision making .

But if you get a test on a piece of paper and it's got a very clear result , it's quite hard not to be influenced by that .

So like , for example , you know , when you get elderly patients , particularly women , and they've had a urine sample sent because it was positive on a urinalysis and it turns out that of an E coli and culture , I think a lot of people struggle to say , like you know , we as a lab have said here's a urine sample that's positive for this organism .

And they end up getting antibiotics even though people are like , well , they're not any symptoms , it's probably asymptomatic bacteria . But then suddenly you're doubting , like , or they just maybe not mentioned symptoms , or they're mildly confused , maybe that's because of the ETI .

And the listener may be wondering why do we care so much if they're getting just a short course of 94 Antone or trimethybrom ? Well , what if they're not ? What if the clebsiella that comes back because XDR ?

That's happened to me a couple of times and I was agonizing over what to treat them with and then I went and asked the team and then asked the patient like wait a second , do you have any symptoms because of this ? Let's see .

Well then it's just a symptomatic bacteria and you know , if you want to treat this , you'll need to change the catheter and we'll need to use a carbapenem or we should do a full episode on catheter associated TIs , because I don't think we should .

So if you you're suspecting some sort of renal disease , but in the context of diagnosing an infection there's not many other circumstances , you should never your your analysis a catheter specimen , because basically you've got plastic in your urinary tract , you're going to have a positive leukocyte estuaries and there's probably going to be bacteria there and those bacteria

will have nitrate reductase , so nitrates will probably be positive as well . Nothing about whether the patient has a UTI or not . So don't never send you know , never base any decision on that and just don't do it . And it happens again and again and again and again . So as a plea , please , please , don't you're in dip catheter specimens of urine

.

But if they've got you know , good going symptoms of UTI , then you're not going to rely on a negative urinalysis , you're going to get urine culture . True enough , like yeah . And the other group is don't you know , an elderly patient ? So particularly women over 65 is the sort of in the guidelines , though maybe 65 is too young .

Now there's a high rate of asymptomatic bacteria and urinalysis can't really tell you the difference between asymptomatic bacteria or bacteria . So you know , sending urinalysis in that situation is not helpful either .

If they have the symptoms we talked about earlier on they've got frequent urinary frequency and dysuria , they're pre test probability . So how likely they are to have a UTI before doing anything else has gone up to 73% already . And if they don't have vaginal discharge , that's up to 90% . So before doing the urinalysis is 90% chance this patient's got UTI .

We can then do the urinalysis . If they've got a positive nitrate or plus or minus Lucas esterase , we're looking at 97% post test probability . So in that situation where you think they probably got UTI , you're already 90% . You do the urinalysis is positive , bang , you're 97% . They've almost certainly got UTI .

But if you've done the urinalysis and not that group and they're negative for nitrates and Lucas esterase , they're still more likely than not to have a UTI . You know they might have a low level of bacteria in the urine . So it's still 71% of patients in that group with negative urinalysis . But they have the really strongly suggested symptoms .

They're still more likely to have a UTI . So you're still going to think I'll give them some antibiotics in the urine culture . So is that actually that useful ? And then the other line of the end of the spectrum , I guess , is a patient that comes in and they don't have urinary frequency and they don't have any dysuria Before you do any tests .

Their pre test probability is sitting down at 23% . And if they've got vaginal discharge then you're down to only 8% of those people have a UTI . So it's much more likely they've got an STI at that point , even if you have a positive nitrite plus or minus . Look at the site esterase . In that group your post test probabilities only 27% .

So one in three people essentially might have a UTI . So it's still more likely not , they don't have a UTI . And also because essentially you've got an STI that can give you a positive look at the esterase and other tests . So is it that useful there either ? And then basically what you want is a test that's going to say yes or no , is it a UTI or not ?

And you can talk about having a good sensitivity and specificity . But really what's in my head important is probabilities and these are like in the lab , in like a gold standard setting in a research paper you can say is a very sensitive test and it's very specific .

But when you have patients and you put into the mix like the actual history and the situation you're in , right there you really have so much information that your analysis rarely is going to say definitely yes or definitely no , you're often just left thinking , well , it could still be or it might not be .

It's maybe like an extra tool in your armamentarium of things that are supportive evidence and things that are not supportive evidence . But then I think the issue of that is that we , as clinicians , are prone to putting more weight onto things that are seen as objective than things that are seen as subjective .

Even though actually symptoms and a lesser extent signs are much more useful in most situations and it's not an expensive test , you know . That's fair to say . Compared to a lot of the other things that we do , it's not expensive .

It's a big automated lab robots and yeah , but lots of conveyor belts is very cool . You know , like just because it's cheap doesn't mean it's good , and just because it's cheap doesn't mean you should use it on everyone .

So if you think about anyone with a catheter all men and all women over the age of 65 , your group that you should only be using in is already un-cathed drives women with mild to moderate symptoms under the age of 65 . That's like not a lot of people . Most of our patients are older than 65 .

But the thing is , most of those patients will have uncomplicated UTI and you're not meant to do your analysis then anyway , you just get them antibiotics into them underway , or actually , but will they resolve without antibiotics ?

There's a lot to say about urine-eutecton infections , about diagnostics , urinary culture , et cetera , and also treatment . When do we use your analysis ? What sort of takeaway from this , James ? How do you become a responsible urine dipper and not be a misprofit ?

Like those groups of patients that we've already said it's not useful and don't do your analysis on them ? That should be a takeaway message Don't urinate , alice . Like after specimen urine .

And in patients where it's really clear , like they've got barn door symptoms , you know there's pus coming , you know , like who knows what it is , but it's like barn door UTI , it's very clear . It's not useful there because even if it's negative , still thing hits UTI .

And in patients where it's like you're really not thinking it's a UTI , again it's not really useful . I think it's probably of some benefit where , like maybe they've got one symptom , like they go about dysuria frequency , you're not particularly certain and you're really on the fence about whether or not to give them treatment or send a urine culture .

I think a positive urinalysis in that situation might just push me slightly towards the side of saying this maybe is a UTI , but if you look at the percentages it's still not certain . So it's not going to give you a definitive answer and it's not something to be relied upon . It's just an extra little bit of you know .

It's the cherry on top of the diagnosis of UTI . It's not really the meat . The meat , the flesh no , that's not the right term the meat of the pudding . I'm looking for analogy here and I can't quite hold on to the phrase cheese of the cheesecake . I think I'm hungry . I'm making food analogies .

I suppose the issue , cal , is what if the patient can't tell you ? You know , old man comes in confused , off legs , not knowing why . Chest X-ray there's maybe a bit of fluid on the bases . You don't know if it's infection . You don't know if it's not . There's no cellulitis . Abdomen's okay . Do they have a UTI ? Do they have a chest infection ?

Why don't I just cover both ? Do you know that kind of case ? Yeah , we've had hundreds of cases like that and do you know ? Do you think the UTI ? Your analysis got any utility there or do you gain ? Do you just think because it's male ? Well , if you look at . So there's a couple of studies . You can meet the patient female , can't you ?

So D'Chamal , d'villetal and D'Fanny Metaal I'm definitely saying them all wrong and they looked at nitrite plus or minus lucasite estuaries or most of them did anyway in the older population , and they quoted a sense of sensitivity and specificity . I don't really interested in that .

So the positive predictive value if you had nitrite plus or minus lucasite estuaries is about 45% . So if you took 100 patients that you thought might have a UTI and they were positive for nitrite plus or minus leukocyte esterase , only 31 to 45% of those actually likely to have a UTI and the negative predictive value is better , so it was 92 to 100% .

So I guess the use case there is if you're thinking I'm not sure what's wrong with this patient and you do your analysis and it's positive , that does not mean they have a UTI and I think that's what is usually interpreted as yeah , but actually , particularly in older patients , often it's going to be something else or they've got asymptomatic bacteria .

The issue is what if the guy doesn't have a UTI that's causing them problems , but you dip it , or the nurse dips it , and they find white cells and blah , blah , blah and they say , oh well , I'm going to send it away . And it comes back and you've got a culture result . And what do you do with that ? Do you treat it ?

You're obviously more likely to treat it now than if you didn't have the results sitting in front of you . And even if it's a plain bog , standard UTI and you can treat it with normal antibiotics , forget multiple resistances . Like I said before , that is increasing the antibiotic burden . That is driving antimicrobial resistance . That is not a good thing .

It's not just related to infectious diseases , like an emergency medicine sort of driven thing , but one of them , I'm pretty sure , was not doing dipsticks in people that don't have symptoms of it and that's the thing .

If you just confine yourself to not doing them in people that don't have urinary tract symptoms already , you would have eliminated at least half of all the urinalysis that are performed in the NHS .

Yeah that's true , and eventually there would be a cost saving associated with it , if not a cost saving of the urinalysis dipsticks themselves , because they are inexpensive , but it does take staff time to run them , that's true .

We spent a lot of this episode basically just crapping all over urinalysis .

We're going to do a UTI episode shortly and we'll mention urinalysis there , but most of our chat will be about urinary culture , because that's probably the main thing that we use and that's not for diagnosis , that's for finding out what the bug is so that we can treat it , because you diagnose it on the symptoms .

There's a link in the podcast show notes . And until next time , I'm James , I'm Kallen . See you there . Now that the episode's done , we hope you learn and had lots of fun .

So go forth and treat people with some of what you now know .

Oh , thanks , man , I should listen back to this and then write it down so that I can see it at the end of every episode , you know there's actually this wonderful technology that we're using , where I could just copy that very tidbit there and then put it into the next episodes , like we do with the music .