Learning Objectives: After reading this article, the physician should be able to:
- Discuss the relevance of hematuria.Explain the use of IV hydration in treating kidney stones.Describe the value of CT scanning for diagnosing and treating nephrolithiasis.
Release Date: March 2006
Emergency physicians frequently evaluate patients who have signs and symptoms of ureterolithiasis. It's a very painful and often dramatic presentation and a time when an emergency physician can look like a pro. You can nail a diagnosis that causes much distress and certainly a whole lot of pain.
Sometimes the diagnosis is easy, with a classic history and readily positive tests. Other times it's more challenging and can be confused with or considered instead of a few conditions that are much more problematic for the patient. Prognosticating a kidney stone and asking the nurse to get the morphine as the writhing, sweating, vomiting patient is wheeled past the nurses' station by EMS impresses everyone. Mistaking an aortic dissection for a stone impresses no one except the hapless clinician and the hospital's risk manager.
Most of us have our routine work-up, and all of us have our particular bias and dogma that we expound to anyone who will listen. As with any medical entity, there are a number of myths and misconceptions that have been promulgated concerning the diagnosis and treatment of ureteral calculi. Some approaches are adhered to religiously despite data that are questionable, nonexistent, unsupportive, or downright contrary.
In this month's column, I will review some of the commonly held clinical beliefs and practices concerning kidney stones in an attempt to evaluate which are truly axioms and which are merely unsubstantiated rumors.
Myth: A plain abdominal radiograph is clinically useful in diagnosing and treating renal calculi. It is a cost-effective screening study.
Fact: Plain films have such a high incidence of false positive and false negative readings that interpreting them for ureterolithiasis is essentially guessing and useless for making firm diagnostic and treatment decisions. If a radiopaque stone has been demonstrated by other means, a plain x-ray may be used to assess progress in selected cases.
Use of Abdominal Flat Plate Radiographs in Patients with Suspected Ureteral Calculi
Zangerle KF, et al
Ann Emerg Med
The authors of this study tested the ability of clinicians to detect and localize ureteral calculi on plain abdominal radiographs. The scout films of intravenous pyelograms (IVPs) of 120 patients undergoing this study for various reasons were presented without a medical history or laboratory data to a staff radiologist and a staff emergency physician. These plain films were blindly read by the two specialists who were instructed to make an interpretation of the presence or absence of a ureteral stone, localize the calculus, and then grade their degree of confidence in that interpretation.
To minimize the chance that the calculus had moved during the time between the flat plate and the IVP performance, the plain radiographs were actually the scout films used for the IVP. Approximately half of the IVPs (51 cases) demonstrated a ureteral calculus (40 cases) or obstruction without definite calculi (11 cases). Of 69 calculi-negative IVPs, 23 were ordered for suspected calculi and 46 were for evaluation of hematuria, tumor, or trauma. For the radiographic interpretation to be considered a true positive, the reader had to assign a “likely” or better confidence level to his reading, and the stone identified on the plain radiograph had to be in the same position as noted on the IVP.
Overall, attempts to utilize the flat plate were dismal. Only half of the readings were correct about the presence or absence of a calculus, but the incidence of true positives was only 33 percent for the emergency physician. The incidence of true negatives was 76 percent and 71 percent for the radiologist and EP respectively. The interpretations of the radiologist and emergency physician agreed with the IVP only 51 percent and 48 percent of the time respectively. The radiologist was correct on 36 percent of the positive and 64 percent of the negative films. The corresponding percentages for the emergency physician were 33 percent and 60 percent respectively. Both readers rarely registered a high confidence or “definite” prediction for the presence of the calculus, a criterion for being called a true positive. For example, of 133 readings by the radiologist, a “definite” prediction was given only eight times, and for five of those, the interpretation was incorrect.
Although the study only used two physicians for x-ray interpretation, and the readers were presented the films without the benefit of history or laboratory data, the authors concluded that the plain abdominal radiograph is highly unreliable in patients with suspected ureteral calculi. This x-ray adds no useful information to the evaluation of suspected kidney stones. The false positive and false negative interpretation rates were high, making the plain film a useless study. An IVP was recommended for definitive diagnosis in all cases of suspected ureteral calculi.
Comments: I reviewed this paper only for medical curiosity's sake. It's difficult for today's house staff to believe that clinicians once relied on such antiquated tests. Today the superiority of the unenhanced spiral CT scan for the evaluation of acute flank pain and for confirmation of suspected renal/ureteral calculi is indisputable. The IVP and flat plate x-ray are now dinosaurs that need to be relegated to the nearest tar pit for the modern emergency clinician who has access to a CT scanner. Although the IVP was generally definitive, it only provided information for a few conditions. A plethora of other important pathology also is seen with the noncontrast CT scan, and it takes only a few seconds, so there is no need to quibble. And most of the time, you don't even need a radiologist — even I can see most renal stones; they light up like bright stars on the CT scan. In addition to placing the stone and measuring it, residual hydronephrosis from a recently passed stone can be detected. It's almost too easy with the CT scan.
Most residents (even radiology residents) have never ordered an IVP, and certainly would have trouble interpreting one in the middle of the night. But I am old enough to remember when the IVP was the gold standard for diagnosing kidney stones. It usually performed well, but the test included IV dye injection, a few hours to complete, and a lot of hemming and hawing and guesswork. Some trauma protocols included a routine injection of dye and a flat plate for all trauma patients going to the OR.
Tradition is difficult to change, and some of my colleagues continue to order plain films (also called flat plate, KUB) as a “screening test” for patients with suspected kidney stones. For some odd reason, most can quote these studies or confirm the shortcomings with their clinical experience, but still order the films anyway. Occasionally an unenlightened radiologist will request this study before agreeing to perform an IVP or CT, even though he will repeat the plain film (AKA scout film) as part of the preliminary evaluation. If a definitive study is scheduled, it's absurd for the EP to order a plain film. Although one can occasionally see a stone on a plain film, it is such an inaccurate study that the procedure should be abandoned. Once a radiopaque stone is confirmed by CT/IVP, occasionally the flat plate may be used on an outpatient basis to follow the movement of that stone, but this is the only reason that I can see for ordering such a low-yield study.
Roth et al (Ann Emerg Med 1985;14:311) conducted a similar study that retrospectively reviewed the IVPs of 206 adult patients who underwent an emergency dye study as part of their evaluation for possible kidney stones. A staff radiologist was asked to interpret a plain film blindly, and the results were compared with the IVP reading. It's noted that the incidence of true positives was again less than 50 percent. The sensitivity and specificity were only 62 percent and 67 percent, respectively. These authors similarly conclude that the sensitivity, specificity, accuracy, and negative predictive value of plain films are low, and the x-ray provides no more diagnostically useful information than the history, physical examination, and urinalysis.
There are many reasons the abdominal plain film is a useless test if it is ordered for renal colic. Although textbooks state that 90 percent of stones are radiopaque, this is a misleading statistic. It is true that most stones will contain radiopaque calcium. Once you know that the stone is present, it may be possible to diagnose it accurately 90 percent of the time with a retrospective review of the flat plate. In my experience, however, the incidence of radiopaque stones that can be seen on the initial plain film and localized to the GU tract does not come close to 90 percent. In the studies reviewed here, less than half of the stones were prospectively called on the initial reading of the plain film; it's more accurate to flip a coin! Even if a stone contains calcium, omnipresent bowel gas, GI contents, feces, and underlying bone, the myriad calcific densities (vascular, nodes) normally seen in the pelvis and abdomen make those little white specks almost impossible to interpret with any accuracy.
Acute Ureteral Colic and Fluid Intake
Scand J Urol Nephrol
The authors of this prospective study assessed the effect of forced IV fluid administration or strict fluid restriction on pain and stone passage rate in 60 patients with acute ureteral colic. During the first six hours after diagnosis, half of the patients were managed with total fluid restriction and the other half were treated with a generous 3 liters of IV fluids (2 liters D5W and 1 liter Ringer's lactate). All patients had an IVP prior to entering the study. The location and diameter of the stone, gender and age of the patient, and duration of the pain prior to evaluation were similar in both groups. The patients graded their pain after six hours of treatment. Meperidine, in less than generous doses, was given for pain control.
The authors concluded that the amount of fluid intake had no practical influence on the degree of pain or clinical outcome in patients with renal colic. The authors note that while the pain of ureteral colic is thought to arise from distension of the renal pelvis and ureter proximal to an obstructing stone, an increase in pressure is not the only factor producing pain. This concept of fluid-induced proximal distension resulting in increased pain has been the basis for recommending fluid restriction. In the presence of ureteral obstruction, however, the majority of administered fluid is filtered by the opposite kidney. Volume expansion does not necessarily produce additional urine flow in the obstructed side.
Most stones spontaneously passed, but the fate of the stone was not related to the fluid intake. The authors conclude that fluid intake has no effect on initial pain relief or spontaneous stone passage in patients with acute ureteral colic.
Comments: It may seem obvious to assume that fluid infused into a peripheral vein ends up in the obstructed ureter, applying proximal pressure to flush out a stone. Some physicians are just as adamant that excess hydration increases pain by increasing distension of the proximal ureter and renal pelvis. Physiological studies would seem to indicate that neither assumption is correct, but this study is one of the few clinical trials that actually examines the concept. The author of this report believes that oral hydration is adequate, and does not advise routine IV fluids. Although this study is limited and has a few methodological flaws (especially the lack of consistent evaluation of urine flow, volumes, and a dynamic pain score), it is one of the few studies of its kind in the literature. I would agree that 3 liters of fluid over six hours is aggressive fluid management, and such a fluid load should have provided an obvious benefit if it were to make a difference.
Myth: Patients with kidney stones should receive high volumes of intravenous fluids to increase urine flow and to flush out stones or patients with kidney stones should be fluid-restricted because excess fluids will distend the proximal ureter and increase pain.
Fact: Neither fluid loading nor fluid restriction affects the incidence of stone passage or degree of pain. Most of the parenteral fluid is shunted to the contralateral kidney and never reaches the involved ureter.
Most patients with kidney stones are vomiting. They come to the hospital very quickly because of the severity of renal colic, and significant dehydration is usually not an issue. Certainly, dehydrated patients should be fluid-loaded prior to receiving any dye study because dehydration is a risk factor for dye-induced nephropathy. Most patients with a stone are young so they can handle fluid. Obviously, excess fluid is detrimental to patients with limited cardiac reserve.
The use of glucagon and other anticholinergic or antispasmodic agents has been suggested to relax the ureter and assist in the passage of the stone. There are no data in the literature to support that this pharmacologic approach is of any value, but more on that next month.
There are simply no data in the literature to support the use of either forced diuresis or fluid restriction as a therapeutic modality in patients with kidney stones. It's probably impossible for the patient, and occasionally the nursing staff, to buy this concept. As a practical matter, I cannot imagine a patient with renal colic not getting routine IV access, if only for pain medication and blood tests, so you won't save the patient an IV stick with the no-fluid mentality. Even though there are no data to suggest that IV fluids either increase pain or force a stone through the ureter, that dripping IV is part of the show business of emergency medicine so it seems like a good idea. Bottom line: Start an IV, but don't drown the patient with saline.
Microscopic Hematuria and Calculus-Related Ureteral Obstruction
Stewart DP, et al
J Emerg Med
The authors of this study examined the correlation between the degree of microscopic hematuria and the degree of IVP-confirmed ureteral obstruction in 160 patients presenting with renal colic. This was a retrospective study that was designed to determine if the number of red cells in a urinalysis could predict the extent of ureteral obstruction. Severe ureteral obstruction, defined as dye extravasation or delayed ureteral filling times of two hours or more, was present in 29 percent of the patients.
The authors found no relationship between the degree of hematuria and the degree of ureteral obstruction. Importantly, seven percent of the patients had severe obstruction and a urinalysis that could be considered normal: fewer than 3 RBC/HPF. Additionally, 11 percent had some obstruction in the presence of a normal urinalysis. The overall incidence of hematuria was 85 percent in patients with nonsevere obstruction and 76 percent in patients with severe obstruction. Clearly, the presence or absence of hematuria was not predictive of the degree of obstruction.
Comments: This article is included because it is one of the few that evaluates the frequently quoted (and rarely referenced) axiom that the absence of red cells in urine sediment correlates with an obstructed ureter. It is also one of the first studies that evaluated the actual incidence of hematuria in patients with proven stones.
Most physicians believe that the vast majority of patients with a kidney stone will have at least microscopic hematuria, and those with complete obstruction will not. I used to tell patients that blood in the urine was a good thing; it meant the stone was moving and might soon pass. That sounded good to me at the time. It's clear that a stone can present with either gross hematuria or a perfectly normal urine, but usually the urinalysis is somewhat abnormal. Other authors have demonstrated an incidence of hematuria as high as 97 percent (Clin Radiol 1980;31:605) or as low as 66 percent (Practitioner 1979;223:387). In my experience, a majority of patients will have at least microscopic hematuria associated with their stones, but you cannot use the test to guide definitive testing and therapy. This study states that hematuria may be absent in as many as one in four patients.
Emergency physicians usually rely on the nurse-run dipstick analysis to define hematuria in those with suspected stones. (J Urol 1984;132:64; Br J Clin Pract 1986;40:269.) Hematuria is probably best defined by microscopic analysis of the urine sediment, however. Greater than 3 to 5 cells per high-power field is generally considered abnormal. The actual quantitation of red cells, however, is dependent on the centrifugation time and volume of supernatant that is used to re-suspend the urine, and finding a few red cells is normal. Most lab techs will evaluate more than one high-power field, but I was always confused by 0 to 5 cells per field. As another practical matter, you simply cannot wait until the lab tech powers up the microscope to order the CT scan or to administer generous amounts of morphine/Dilaudid. And, of course, the microscope cannot differentiate between blood from the ureter and blood clandestinely added to the urine by the precocious drug seeker.
Most physicians use the urine dipstick as the initial method to test urine for blood, and it's the test that often prompts either a narcotic administration or a CT request. (Clin Nephr 1994;41:167.) Overall, the urine dipstick is quite accurate in identifying minute quantities of blood. It can be positive from myoglobin from soft tissue trauma or skeletal muscle injury, and occasionally microscopic confirmation is necessary. I occasionally see a positive dipstick test for blood in the ED, and do not have it confirmed by microscopy. A false negative test is not usually an issue, although dipsticks used to lose their ability to detect small amounts of blood if stored in the open air for prolonged periods. (Am J Clin Path 1991:96:398.) Some dietary peroxidases can give a false positive dipstick test for blood. Excessive oral vitamin C use and a very acidic urine may give a false negative dipstick analysis.
I am not sure whether a formal microscopic urinalysis should always be done in patients with suspected stones to rule out possible infection. Most of us do not order it routinely. A minuscule number of red blood cells, enough to qualify for a “negative” microscopy, can turn a dipstick positive. A healthy patient who gets sudden flank pain and has no fever or dysuria does not have a UTI. As the stone gets to the UV junction, however, it may cause frequency, urgency, or dysuria suggestive of cystitis. Although significant hematuria is rare following careful bladder catheterization, a freely voided urine specimen is always preferred. (Ann Emerg Med 1987;16:550.) Always ask women about menstruation and get a catheter specimen if it is imperative to confirm hematuria, a rare mandate.
Myth: Because almost all patients with kidney stones will have at least microscopic hematuria, a urinalysis for blood is a good screening test to use for deciding further testing. If hematuria is absent, complete ureteral obstruction is likely.
Fact: Up to one in four patients with documented stones will not exhibit hematuria. The magnitude of hematuria does not correlate with the degree of obstruction.
At the risk of being termed a medical heretic, I think it's rather obvious that the presence or absence of hematuria is of limited value in the assessment of patients with acute flank pain or suspected stones. (J Urol 1999;162:685.) A study by Luchs et al (Urology 2002;59:839) confirmed the somewhat disappointing utility of microscopic hematuria testing (not simply dipstick testing) in 950 patients with suspected renal colic. Using 10 RBC/mL or greater as a definition of hematuria, the sensitivity and specificity of hematuria for ureterolithiasis were 81% and 51% respectively when a stone or recent evidence of a passed stone was diagnosed by a noncontrast helical CT scan. The association between hematuria and renal colic was high, but it was not close to being totally reliable. The positive and negative predictive values of hematuria for renal colic were 72% and 65% respectively. The urinalysis was negative for hematuria in 17 percent of patients with stones measuring 5 mm or less. Importantly, hematuria was seen in 52 percent of patients with a completely negative CT scan or an alternate diagnosis. Important alternative diagnoses are aortic dissection/aneurysm, cancer, tuberculosis, endocarditis, appendicitis, sickle trait, UTI, and renal vasculature problems.
Because hematuria testing is not a reliable predictor or negator of ureterolithiasis, it cannot be used as a reliable screening test. I have a difficult time buying this concept, and in my experience, hematuria is much more common than reported in these articles. I guess I will have to order more CT scans in patients with lower abdominal/flank pain who have no hematuria.
Finally, there is some debate even among my colleagues on when you can clinically diagnose a kidney stone and forgo the CT scan. If the patient has a history of multiple stones, typical pain, and hematuria, some will eschew the scan and give time and analgesia a chance to work. There is literature to support this approach. (J Gen Intern Med 1993;8:57.) I take a strong contrarian view to this tactic. Most experienced EPs have seen more than one case that appeared to be classic renal colic, but 10 hours later was diagnosed as another condition that mimicked a stone (and there are many).
I have the mindset always to confirm the diagnosis with a CT scan. That settles the issue, may find other problems, measures the stone size for prognostic purposes, and may even find other stones that need to be followed more closely. In reality, I do not believe that any clinical evaluation is sensitive or specific enough to hang your hat on the diagnosis of a kidney stone. A past history of a stone, flank tenderness, colicky nature, vomiting, and dysuria may be helpful but not diagnostic enough for me to countermand a CT scan. Before I push megadoses of morphine or discharge a patient with oral narcotics and outpatient follow-up, I want to be sure what I am dealing with.
Selected Caveats about Kidney Stones
- Ureterolithiasis is a younger person's disease (20–50 years of age).
- Question the diagnosis of a first kidney stone in older patients; it is more likely another process.
- Stones are most common in Caucasian males.
- Patients with multiple recurrent stones may have their first episode in their 20s.
- The position of a stone in the ureter has some correlation with its position in the urinary tract.
- Proximal stones produce flank pain, while lower stones produce lower abdominal pain.
- Stones at the ureterovesicle junction can produce frequency, urgency, and dysuria similar to cystitis.
- Lower stones can produce pain in the groin, labia, and testicles.
- Peritoneal signs and severe localized abdominal tenderness are uncommon in patients with a kidney stone.
- Obstruction of a ureter initially will not produce an elevated BUN/creatinine.
- Stones less than 5 mm in size will usually pass spontaneously (days to weeks).
- Urinary tract infection in the presence of nephrolithiasis is a potentially serious infectious process in the making.
- Catheterizing the bladder usually does not produce enough trauma to explain significant hematuria (greater than 10 RBC/HPF). Look for another cause.
- The best initial test to evaluate for a kidney stone is the noncontrast CT scan.
- Indications for admission of patients with a known stone are uncontrolled pain, vomiting with inability to retain fluids/analgesics, significant infection, and stones unlikely to pass due to size.
Conditions That May Mimic a Kidney Stone
- Acute appendicitis
- Ovarian cyst, torsion
- Aortic dissection
- Epiploic appendagitis
- Inflammatory bowel disease
- Renal vascular disease (infarction, venous thrombosis, vasculitis)
- Pelvic inflammatory disease
- Ectopic pregnancy
Most cases of ureterolithiasis can be sorted out by clinical parameters and confirmed with a noncontrast CT scan of the abdomen and pelvis. The conditions above may mimic a stone, and must be sorted out without additional testing. Because a confirmatory test is usually ordered, even in classic cases, the diagnoses above should be readily clarified. Most of these conditions can be investigated without the use of oral/IV contrast.
Readers are invited to ask specific questions and offer personal experiences, comments, or observations on InFocus topics. Literature references are appreciated. Pertinent responses will be published in a future issue. Please send comments to firstname.lastname@example.org. Dr. Roberts requests feedback on this month's column, especially personal experiences with successes, failures, and technique.
The reader comments below are in response to Dr. Roberts' December and January columns, “The Search for the Holy Grail of Medical Documentation, or To Template or Not to Template” and “ED Discharge Instructions: Another Chance to Help Patients and Prove Your Worth.” (2005;27:17; 2006;28:17.)
Dear Dr. Roberts: As a medical/legal case reviewer, I hope that your columns on physician charting will be taken to heart by every emergency physician. Most of the hundreds of emergency physician charts I have reviewed are woefully deficient. The failure to document pertinent negatives is especially problematic because it then becomes difficult to rebut the plaintiff's contention that the symptom or sign in question was present, and would have led to the correct diagnosis and treatment.
I agree that the template model is likely best, but I would offer two suggestions. Templates often cram many items into a small space so it is important to take the time to be precise in one's circling and slashing out. The benefits of the template are lost when it is impossible to tell what the physician indicated.
Another pitfall is to get carried away with slashing through everything. I just saw an extreme example of this in a patient with animal bites to his hands. The physician documented a full review of systems for 11 different systems and an examination of seven separate systems while not documenting the necessary components of the examination of this patient's hands! More systems are not better, at least for purposes of malpractice, if they are irrelevant.
— David W. Shapiro, MD, JD, Palo Alto, CA
Dr. Shapiro is the editor of the Professional Liability Newsletter.
Dr. Roberts responds: I agree. Dr. Shapiro must have had the frustration of trying to defend a good clinician who was a poor charter. Fortunately for the patient, most physicians spend more time with the medical aspects of the encounter than the paperwork. My comments were meant partially to address medicolegal issues, but good charting generally equates to good medical care. It's important to document pertinent negatives, but most physicians fall short of this suggestion, and there seems to be no definitive standard of care. Charting only positive findings is still very mainstream, and will likely continue until you are the one getting sued. I don't agree with the “if it wasn't written, it wasn't done” concept, but omitting pertinent negatives is filet mignon to a plaintiff's attorney.
On the flip side, no one always believes that you did it just because you wrote it down. You can easily go overboard, get bogged down, and become too paranoid to discharge/admit anyone without five pages of useless information. Charting criticisms are just one of many examples where physicians feel that sometimes they just can't win, no matter how hard they try, a very discouraging concept and one that has caused more than one superb doctor to leave the profession.