The practical diagnostic value of fecal analysis in the evaluation of patients with chronic nonbloody diarrhea is controversial. It is possible that variations in its value depend on how it is done and how the results are interpreted rather than on its intrinsic value. In the authors’ city, stool analysis has been made easily accessible, with a commitment to quality assurance and interpretation. To evaluate its practical value, the results of stool analysis obtained on stool specimens submitted by gastroenterologists were retrospectively reviewed. The results indicate that stool analysis has substantial practical diagnostic value in patients with chronic diarrhea.
- •
The practical diagnostic value of fecal analysis in the evaluation of patients with idiopathic chronic non-bloody diarrhea is controversial due to questions related to its unpleasant nature, quality control, interpretation of results, and intrinsic value.
- •
To evaluate its diagnostic value we retrospectively reviewed the results obtained on 158 stool specimens that were submitted by orders of practicing gastroenterologists during an 18 month period.
- •
A specific cause of chronic diarrhea was identified in 8% of cases; in addition, steatorrhea was found in 28%, and probable carbohydrate malabsorption without steatorrhea in 18% of cases.
- •
Positive findings were almost as frequent in patients with stool weights less than 200 g/d as in patients with stool weights greater than 200 g/d.
- •
In the authors’ opinion, these results indicate that stool analysis has substantial practical diagnostic value in patients with idiopathic chronic diarrhea.
Introduction
In a substantial fraction of patients with chronic nonbloody diarrhea, the cause is not apparent from the history, physical examination, routine laboratory data, or from a search for an infectious organism. In the United States, such patients often undergo exhaustive diagnostic studies, including colonoscopy with mucosal biopsies; upper endoscopy with duodenal biopsy; computed tomography (CT) scan of the abdomen and pelvis; small bowel evaluation by camera, endoscopy, or CT; octreotide scan; and the measurement of plasma peptides. Additional tests can include celiac disease antibody panel, a search for unusual pathogenic microorganisms, culture of jejunal fluid for anaerobic bacterial overgrowth, measurement of serotonin metabolites, and measurement of fecal bile acids. However, in many patients, a specific cause cannot be identified even after they have been subjected to most or all of these tests. The causes of diarrhea that escape detection by these methods include inadvertent or surreptitious laxative ingestion; excess consumption of poorly absorbed carbohydrates; several causes of undetected malabsorption; undiagnosed or undiagnosable infections; alterations in colonic bacterial flora ; autonomic neuropathy causing rapid transit; thyrotoxicosis and other endocrinopathies; immune reactions to food; adverse reaction to medications, food additives, or supplements; and anatomic changes in the pelvic floor or anal sphincter that produce fecal incontinence.
The fraction of patients with idiopathic chronic diarrhea who remain undiagnosed is highly variable depending on criteria for patient selection. The undiagnosed fraction will also vary depending on the diagnostic definitions applied. For example, some physicians make a diagnosis of irritable bowel syndrome or functional diarrhea when preliminary tests are negative and stool weight is estimated to be low or measured to be less than 200 g/d. The authors do not do this because they know of no evidence showing that stool weight less than 200 g/d excludes organic causes of diarrhea. Thus, the undiagnosed fraction according to the authors’ definitions would be higher.
In addition to the fact that current strategies often fail to reveal a specific cause of idiopathic chronic diarrhea, the costs of diagnostic tests and procedures are considerable. Therefore, in the opinion of some clinical investigators in the United States, the early phase of evaluation should include a stool analysis that measures fecal weight, fecal fat output, and fecal fluid electrolyte concentrations. Such analysis can classify chronic diarrhea into different pathophysiological categories, which helps select diagnostic procedures that are most likely to reveal the underlying cause and reduce the use of unnecessary procedures. In addition, stool analysis can identify patients who are surreptitiously or inadvertently ingesting osmotic laxatives.
Although the logic for using comprehensive stool analysis in patients with chronic diarrhea seems sound, such testing is not readily accessible in most clinical centers in the United States. Various components of the analysis must be ordered individually from reference laboratories, and the results are reported without an overall interpretation. As a result, comprehensive stool analysis is rarely used. This situation may not be a big loss because in the United Kingdom, the measurement of fecal fat and fecal fluid electrolytes is considered to have little practical value in patients with chronic diarrhea. This negative opinion seems to be based on the inconvenience and unpleasant nature of collecting and analyzing stool specimens, lack of quality control, and problems related to the interpretation of the results. It is possible that this negative view was also influenced by data showing that, in general practice, stool analysis is inherently unable to provide results that are diagnostically useful; if such data exist, the authors are not aware of it.
The authors accept the view that if analytical procedures are not accurately performed, and if results are not properly interpreted, stool analysis would have little practical value in detecting the cause of idiopathic chronic diarrhea. However, the authors think there is another reasonable question to ask. If stool analysis were accurately performed and interpreted, would it then have practical value?
Present study
About 25 years ago, Baylor University Medical Center established a laboratory to facilitate the evaluation of patients with chronic diarrhea. The laboratory is under the administrative direction of the gastroenterology department, and quality assurance is monitored by a third-party accreditation agency, Commission of Office Laboratory Accreditation (COLA). One of this laboratory’s missions is to perform and interpret a defined panel of analytical tests on stool specimens collected from patients with chronic diarrhea. Many practicing gastroenterologists in the Dallas area now use this laboratory in the evaluation of their private patients. A recent survey of these gastroenterologists revealed that most consider stool analysis as completing a diagnostic workup rather than as a guide for cost-effective utilization of other diagnostic procedures. They usually request stool analysis after endoscopic procedures and biopsies have failed to reveal a diagnosis. These gastroenterologists think that comprehensive stool analysis is extremely useful to them and to their patients.
In an effort to gain insight into the practical diagnostic value of comprehensive stool analysis in adult patients with chronic diarrhea, the authors retrospectively studied results obtained on 158 stool specimens that were submitted by orders of practicing gastroenterologists during the 18-month period between January 1, 2010 and June 30, 2011. This study was performed without knowledge of other diagnostic tests that had been performed or the cause of diarrhea that may have been ultimately discovered. It was approved by the institutional review board of Baylor Research Institute.
Present study
About 25 years ago, Baylor University Medical Center established a laboratory to facilitate the evaluation of patients with chronic diarrhea. The laboratory is under the administrative direction of the gastroenterology department, and quality assurance is monitored by a third-party accreditation agency, Commission of Office Laboratory Accreditation (COLA). One of this laboratory’s missions is to perform and interpret a defined panel of analytical tests on stool specimens collected from patients with chronic diarrhea. Many practicing gastroenterologists in the Dallas area now use this laboratory in the evaluation of their private patients. A recent survey of these gastroenterologists revealed that most consider stool analysis as completing a diagnostic workup rather than as a guide for cost-effective utilization of other diagnostic procedures. They usually request stool analysis after endoscopic procedures and biopsies have failed to reveal a diagnosis. These gastroenterologists think that comprehensive stool analysis is extremely useful to them and to their patients.
In an effort to gain insight into the practical diagnostic value of comprehensive stool analysis in adult patients with chronic diarrhea, the authors retrospectively studied results obtained on 158 stool specimens that were submitted by orders of practicing gastroenterologists during the 18-month period between January 1, 2010 and June 30, 2011. This study was performed without knowledge of other diagnostic tests that had been performed or the cause of diarrhea that may have been ultimately discovered. It was approved by the institutional review board of Baylor Research Institute.
Stool collection, methods of analysis, and interpretation
When stool collections were done at home, patients were instructed to collect stools quantitatively for a defined period of time and to continue to eat their regular diet during the collection period. Along with verbal and written instructions, they were provided with a Styrofoam (Dow Chemical Company, Midland, Michigan) cooler, ice packs, a collection device, and containers for storing stool during the collection period. Using the provided equipment, the temperature in the cooler ranges from 4°C to 10°C. The time of the stool collection was 24, 48, or 72 hours, as specified by the referring physician. During the collection period, patients were instructed to record food intake, the time of each bowel movement, and note if any stool was not collected. Patients were asked to deliver their collected stool specimen to the laboratory as soon as possible after the timed collection was complete. When stools were collected in the hospital, the same protocol was used except that stools specimens were stored in a portable refrigerator and, on completion, they were returned to the laboratory by the nursing staff.
Specimens from 32 patients were not included in the present study of 158 patients for the following reasons: 11 reported incomplete stool collections; for 8, the referring physician noted previous intestinal surgery, which would probably be the cause of their diarrhea; in 5 specimens, the technician detected urine contamination; 3 collected their specimen while fasting; 3 had missing or incomplete data; and in 2, specimens were collected on more than one occasion and only the first was included.
On receipt of the specimen, stool consistency was graded. Stools were then weighed and homogenized. A sample of the homogenized stool was analyzed for fat concentration by the van de Kamer method. Fecal fat output (grams per day) was calculated by multiplying the fecal fat concentration by stool weight. When fat output was greater than 7.0 g/d, the stool was analyzed for chymotrypsin activity. The homogenized stool specimen was also analyzed for excess fecal neutrophils by the Wampole Leuko Ez Vue method for lactoferrin and for occult blood by Hemoccult II. Another sample of the homogenized stool was centrifuged (25,000 rpm for 1 hour), and the fecal supernatant was analyzed for sodium and potassium by Jenway PFP7/C flame photometer (Essex, England), for chloride by Labconco chloride analyzer (Labconco, Kansas City, Missouri), for bicarbonate by Corning 965 carbon dioxide analyzer (Essex, England), for pH by Radiometer PHM 82 Standard pH meter (Radiometer America Inc, Westlake, Ohio), and for osmolality by freezing point depression (Micro-Osmometer model 3MO; Advanced Instruments, Norwood, Massachusetts).
The results were written on a report form and then interpreted by one of the authors (JSF) without knowledge of the patient’s medical history or other diagnostic tests that had been performed. The interpreter sometimes recommended further analytical studies on the stool specimens. The report form was faxed to the referring physician within a few days of receipt of the specimen. The form the authors use, as well as the results and interpretation of one highly selected and dramatic case from the present series of patients, are shown in Fig. 1 .
In this article, the authors do not present or discuss results on occult blood, chymotrypsin activity, or lactoferrin. Their inclusion would have required many additional subdivisions that would obscure the primary objective of evaluating fecal weight, fecal fat, and fecal fluid electrolyte concentrations.
Definitions
Objective Evidence of Diarrhea
Stool consistency was graded according to the following scale: (1) formed, firm; (2) formed, soft; (3) unformed, soft; (4) unformed, mushy; (5) runny; (6) watery. Grades 1 and 2 were considered normal consistency and grades 3 to 6 were considered abnormal.
An abnormal increase in bowel movement frequency was defined as more than 3 bowel movements per day. Hyperdefecation was defined as 4 or more bowel movements per day when stool weight was less than 200 g/d. The authors do not use the term hyperdefecation when stool weight exceeds 200 g/d. Objective evidence of diarrhea was considered to be present when 1 or more of the following criteria were met: (1) consistency graded greater than or equal to 3, as described earlier; (2) stool weight greater than 200 g/d; and (3) average bowel movement frequency greater than 3 times per day.
Steatorrhea
The upper limit for fecal fat output in healthy patients ingesting their normal diet is approximately 7 g/d when measured by the van de Kamer method. However, healthy patients with diarrhea induced by laxatives can have fecal fat outputs as high as 14 g/d, despite maintaining their normal diets. Thus, even when intake, digestion, and absorption of dietary fat are normal, diarrhea per se can cause a secondary steatorrhea, with fecal fat outputs as high as 14 g/d. Therefore, in patients with diarrhea, fecal fat output between 7 and 14 g/d will have a low specificity for identifying primary defects in fat digestion or absorption. Based on the progressive increase in fecal fat output as stool weight increases, the authors defined fat malabsorption in patients with chronic diarrhea as follows: stool weight up to 400 g/d, fecal fat output greater than 7 g/d; stool weight 400 to 1000 g/d, fecal fat output greater than 11 g/d; stool weight greater than 1000 g/d fecal fat output greater than 14 g/d. The authors have no data on fecal fat output by healthy patients with laxative-induced diarrhea who have stool weights greater than 1800 g/d. Nevertheless, in the authors’ laboratory, patients with extreme diarrhea (2000–4000 g/d) are considered to have fat malabsorption only if their fecal fat output exceeds 16 g/d.
Dietary fat intake in the patients was unknown. Patients were not asked to change their diet to standardize fat intake because it takes at least 1 week to develop a new steady state and because actual fat intake would remain unknown. If fat intake in a patient was low, fecal fat output might be within normal limits in a patient who has fat malabsorption. However, in patients with fat malabsorption caused by nontropical sprue, Comfort and colleagues found that fecal fat output on a diet containing about 50 g/d of fat exceeded the upper limit of normal obtained with a test diet that contained approximately twice as much fat. If fat intake was very high, it might cause spurious steatorrhea, although the magnitude of excess fecal fat would probably be trivial because when normal people ingested 30, 60, 100, and 200 g/d of fat, average fecal fat output was 3.2, 4.0, 4.8, and 7.8 g/d, respectively. Based on these considerations, the authors think fecal fat output has diagnostic value even when fecal fat intake is unknown. Obviously, the interpretation of fecal fat output would be more accurate if dietary fat were known and taken into account. The accuracy of daily fecal fat measurement is also limited by the short collection periods (1–3 days). The errors induced by short collection periods are probably less in patients with either voluminous diarrhea or increased bowel movement frequency than in healthy subjects without diarrhea.
Distinguishing Secretory from Osmotic Diarrhea
As explained elsewhere, fluid in the colon and rectum is equilibrated with extracellular fluid within the body and its osmolality is approximately 290 mOsm/kg. The traditional way in which electrolyte and nonelectrolyte solute composition of fecal fluid is expressed is by calculation of the osmotic gap.
The sum of sodium and potassium concentrations is multiplied by 2 to account for associated, mostly monovalent anions; the result approximates the total concentration of electrolytes in fecal fluid. Even though measured osmolality is not used to calculate the osmotic gap, it is always measured to detect the dilution of the stool with extraneous water (measured as a fecal fluid osmolality less than 290 mOsm/kg).
Secretory diarrhea is caused by reduced absorption or increased intestinal secretion of electrolytes and water. In this type of diarrhea, the sum of [Na] and [K] is close to 140 mEq/L, and the osmotic gap is near zero. In osmotic diarrhea, active absorption and secretion of electrolytes are normal, but some extraneous osmotically active solute exists in intestinal fluid that prevents water absorption. In this form of diarrhea, the osmolality of fecal fluid is, to a substantial extent, made up of nonelectrolytes, and the concentration of monovalent electrolytes is correspondingly low; therefore, the osmotic gap is large.
The cutoff points for osmotic gap that distinguish secretory from osmotic diarrhea cannot be derived from a study of patients with chronic diarrhea because in most patients there is no independent way of knowing the mechanisms of their diarrhea. However, electrolyte concentrations of fecal fluid in healthy subjects with diarrhea induced by substances that produce secretory or osmotic diarrhea have been measured. From these results, cutoff values for osmotic gap, chloride concentration, and pH were found that can (1) distinguish between osmotic and secretory diarrhea and (2) suggest when magnesium, polyethylene glycol, sulfate, or phosphate is being used to produce factitious osmotic diarrhea. These results have been described in detail elsewhere.
Carbohydrate Malabsorption
Unlike fat malabsorption, carbohydrate malabsorption cannot be measured simply by analyzing stools for carbohydrate content because colonic bacteria metabolize some or all of the carbohydrate that is unabsorbed by the small intestine. This fermentation takes place within the lumen of the colon and also to some extent in the collection container after stool is passed. For this reason, the amount of carbohydrate excreted in stool does not accurately reflect the amount of carbohydrate that was not absorbed. For research purposes, it is possible to measure carbohydrates and organic acids in stool and to quantitate total carbohydrate equivalents appearing in stool, but these methods are too time consuming to incorporate into a routine clinical stool analysis panel. Stool analysis to detect carbohydrate malabsorption for clinical purposes has, therefore, depended on 2 measurements: fecal electrolyte concentrations, which are not altered by bacterial fermentation after stools have been passed, and the pH of fecal fluid. Unlike fat malabsorption, wherein output is quantitatively measured, carbohydrate malabsorption is defined only as present or absent; the accuracy of the result is not dependent on the quantitative collection of stools.
Diarrhea induced in healthy subjects by the ingestion of lactulose or sorbitol 4 times daily was almost always associated with a fecal fluid osmotic gap between 75 to 240 mOsm/kg. This wide range of elevated gaps is mainly caused by differences in the degree to which unabsorbed sugars in colonic fluid are converted by colonic bacteria to organic acids. The colon reabsorbs a fraction of these organic acids, but unabsorbed organic acids in colonic fluid form salts by combining with sodium and potassium. When fecal fluid contains high concentrations of sodium and potassium salts of organic acids, the sodium and potassium concentrations in fecal fluid increase and the osmotic gap is low. On the other hand, when colonic fluid contains high concentrations of monosaccharides or disaccharides, the osmotic gap is high. In addition to an osmotic gap greater than 75 mOsm/kg, almost all subjects ingesting lactulose or sorbitol had a fecal fluid pH less than 5.5. In contrast, other causes of experimental diarrhea in healthy subjects were almost never associated with a fecal fluid pH less than 5.5.
Based on the results with lactulose and sorbitol, it might be expected that diarrhea produced by carbohydrate malabsorption is always associated with both a large osmotic gap and a low fecal pH. However, when patients with chronic diarrhea caused by known carbohydrate malabsorption (diagnosed by the measurement of fecal water obligated by reducing sugars, organic acids, and associated cations) were studied, only 2 of 6 patients had a fecal fluid pH less than 5.5. This finding suggests that there may be a pH discrepancy between healthy subjects ingesting lactulose and sorbitol compared to patients with disorders that impair absorption of normal dietary carbohydrates.
To obtain further information on malabsorption of physiologic dietary carbohydrates, for the past 3 years the authors have studied fecal fluid from patients who develop diarrhea while undergoing a breath hydrogen test to detect lactose malabsorption. (None of the patients undergoing breath tests were part of the 158 patients with chronic diarrhea that are the main focus of this article.) The main goal was to study concordance between fecal fluid osmotic gap and pH criteria that are used for the identification of carbohydrate malabsorption. In addition to the usual tests in the stool analysis panel, the concentrations of total carbohydrate and organic acids in fecal fluid were also measured.
Table 1 shows results in 16 patients who developed diarrhea during a 3-hour period after ingesting 50 g of lactose and whose fecal fluid had either an osmotic gap greater than or equal to 75 mOsm/kg or a pH less than 5.5. Fourteen of these patients had a positive lactose breath test, whereas 2 did not. Fecal fluid osmotic gap was greater than 75 mOsm/kg in all 16 patients, whereas fecal fluid pH was less than 5.5 in only 3 patients. No patient had a low pH without an associated increased osmotic gap. Fourteen had an abnormally high fecal fluid concentration of carbohydrate compared to normal subjects with diarrhea induced by polyethylene glycol (PEG). In contrast, the concentration of organic acids in fecal fluid was elevated in only 4 of 15 patients. There was no statistically significant correlation between organic acid concentration and pH ( R = 0.33, P = .237).
| Lactose Breath Test a | Stool Weight Within 3 h After Lactose Load (g) | Consistency | Concentrations (mEq/L) | Osmotic Gap (mOsm/kg) | CHO (g/L) | Organic Acids (mEq/L) | pH | |||
|---|---|---|---|---|---|---|---|---|---|---|
| H + (ppm) | ||||||||||
| Baseline | Peak | Na | K | Cl | ||||||
| 2 | 21 | 355 | Runny | 49 | 9 | 32 | 174 | 20.9 | 24 | 7.20 |
| 3 | 92 | 54 (FI) | Runny | 32 | 32 | 4 | 162 | 1.9 | 66 | 5.81 |
| 0 | 246 | 171 | Runny | 32 | 33 | 16 | 160 | 6.3 | 106 | 5.46 |
| 0 | 44 | 97 | Soft | 21 | 49 | 15 | 150 | 10.8 | NA | 5.65 |
| 3 | 192 | 48 | Runny | 39 | 32 | 64 | 148 | 26.2 | 120 | 5.22 |
| 0 | 83 | 319 | Runny | 38 | 35 | 15 | 144 | 31.3 | 123 | 5.60 |
| 2 | 98 | 103 | Runny | 41 | 33 | 30 | 142 | 9.0 | 124 | 5.99 |
| 11 | >400 | 60 | Soft | 8 | 67 | 16 | 140 | 1.9 | 118 | 6.87 |
| 2 | 310 | 365 | Watery | 31 | 49 | 23 | 130 | 21.6 | 89 | 5.81 |
| 2 | 82 | 517 | Watery | 67 | 16 | 54 | 124 | 9.9 | 28 | 6.91 |
| 12 | 157 | 261 | Runny | 58 | 31 | 34 | 112 | 16.5 | 149 | 5.25 |
| 1 | 186 | 400 | Runny | 24 | 66 | 16 | 110 | 18.0 | 126 | 6.05 |
| 15 | 208 | 220 | Runny | 36 | 56 | 15 | 106 | 6.2 | 402 | 5.75 |
| 14 | 248 | 359 | Runny | 40 | 57 | 25 | 96 | 8.4 | 155 | 6.30 |
| 17 | 22 | 112 | Soft | 34 | 72 | 8 | 78 | 7.2 | 212 | 6.12 |
| 19 | >400 | 78 | Runny | 64 | 43 | 45 | 76 | 6.4 | 129 | 6.88 |
Related posts:
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
