EDITORIAL

The Bristol scale - a useful system to assess stool form?

Escala de Bristol: ¿un sistema útil para valorar la forma de las heces?

M. Mínguez Pérez and A. Benages Martínez

Service of Digestive Diseases. Hospital Clínico Universitario. Valencia, Spain

In clinical practice, difficulties in assessing stool characteristics (consistency, form, smell, color, etc.) are common during history taking. This is not only due to patient or caregiver squeamishness regarding attentive fecal inspection for each bowel movement, but also to a number of factors including the variability of stool form and consistency among individuals or in one individual over time, and changes in stool form and consistency during one bowel movement (1,2), since some individuals commonly evacuate hard, ball-shaped feces early during defecation followed by soft or even fluid stools subsequently. Furthermore, hard-to-homogenize variables also apply, including variously designed toilet pans that on occasion distort or completely hinder an assessment of fecal characteristics. An easy-to-use, accessible system to quantitize stool consistency and form would be an important asset, particularly one allowing an understanding of the relationship between fecal characteristics and defecation-related patient complaints. Other goals to consider include an analysis of population-related variability according to food types, age, sex, drugs, and lifestyle.

Stool collection and subsequent fecal testing in a laboratory is virtually an impossible thing to do in daily practice or epidemiological studies, this method remaining a restricted option for research with small sample sizes (1,2). An objective, validated, self-administered, easily quantifiable recording system would considerably speed up data collection by physicians. Two descriptive systems have been published that allow to record fecal form and consistency (2,3). The scale by Davies et al. (1) establishes 8 descriptive options - from type-1 or watery feces to type-8 or hard, fragmented, goat-like stools; type 5 corresponds to smooth cylindrical feces. This scale has not been validated but proved useful to demonstrate - in a reduced sample (n = 51) assessed for food ingestion - that defecatory frequency, and stool weight and rheology made up the best clinical marker for bowel transit time (as measured with radio-opaque substances). Thus, stool shape corresponded to either a fast or slow transit (type 1 was related to fast transit, type 8 was associated with slow transit). The statistical power of this association allowed these authors to hypothesize that simply inspecting stool form could be useful to estimate bowel transit time in clinical practice. Using this same scale, Aichbichler et al. (2) attempted to evaluate differences in stool characteristics between constipated subjects (n = 20) and a control group (n = 20); fecal consistency was analyzed in a laboratory for stools collected over 7 days. These authors only found significant differences in weekly mean weight, which was lower for constipated patients, but not in stool shape using a descriptive scale.

The so-called "Bristol scale" was developed and validated in Bristol by Heaton et al. (3) in order to descriptively and graphically assess 7 stool types according to form and consistency. The relevance of this scale is that it shows the patient drawings illustrating stool shapes together with precise descriptions regarding form and consistency, and using easily recognizable examples (for instance, in type 1, by a color illustration of feces as separate balls, a legend explains: "Hard, separate balls. Like nuts"). The patient has only to select the type that, according to the drawing and description, more closely resembles his or her own stools. The scale is structured from 1 to 7 according to form and consistency, from the hardest (type 1) to the fluid kind (type 7). The method used for scale validation is difficult to assess, as findings were only reported as an abstract (3).

Bristol has been home to the one study analyzing stool form and consistency in the general population (838 males and 1059 females) (4). This study shows that type 4 ("smooth, soft, long, sausage-like feces") is most common (for both genders), whereas hard stools (types 1 and 2) predominate in women (25.3 vs. 17.1% in males), and soft-fluid stools (types 5 and 6) are more common in males (11,9 vs. 8%). In addition, a majority of the population reports that defecation is normal (with no urgency or effort or rectal tenesmus) for type-3 and type-4 feces, while mushy, fluid stools are associated with defecatory urgency in 80% of cases.

This scale has shown that fecal shape correlates to total bowel transit time as measured with scintigraphy or radio-opaque markers (3,5-7), both in patients with irritable bowel syndrome (6) and healthy subjects (5,7); thus, types 1, 2 and 3 correlate with a slow transit, and types 6 and 7 correlate with a fast transit. In 1997 Lewis and Heaton (8) demonstrated in healthy volunteers a significant correlation between Bristol scale values and bowel transit time, both under baseline conditions and after laxative or constipative drug administration; that is, the Bristol scale was highly sensitive to drug-induced bowel transit changes. From the results of these studies Haeton et al. (6) concluded that the Bristol scale was a very useful tool for clinical practice, epidemiological studies, and clinical trials, as it easily and with no radiation allowed to rapidly differentiate individuals with a fast transit time (loose stools) from those with a slow transit time (hard stools).

Endorsing this theory, Choung et al. (9) performed in 2007 a population-based survey through the mail in Olmsted County (Minnesota) using the Bristol scale; the scale was sent to 4,196 selected individuals (54% responded), and three transit time groups were established according to the responses obtained: a) slow transit time (type-1 and -2 feces in the Bristol scale); b) normal transit time (types 3, 4 and 5); and c) fast transit time (types 6 and 7). According to this study one in 5 individuals has a slow transit time, and 1 in 12 show a fast transit time.

Bristol scale simplicity and the results obtained from a number of studies has progressively incorporated this tool into clinical practice for the assessment of patients with irritable bowel syndrome (10-12), HIV-related diarrhea (13) and fecal incontinence (14), among other things. Similarly, this is the one scale currently including fecal shapes as recommended by consensus groups on data collection for functional bowel disease (15).

Parés et al. (16) report in this issue an adapted, validated Spanish version of the Bristol scale. This study is interesting because it validates with an appropriate method a questionnaire in use for over 20 years now, whose original validation remains unknown, and also because of its adaptation to the Spanish culture. The study, performed in healthcare professionals (79 physicians and 79 nurses) and patients (78), demonstrates that consistency between stool form descriptions and their accompanying illustrations is very high for all subtypes in all analyzed populations, except for type 5, which scores very low among nurses (20%) and low among patients (40%). These results prompt a redefinition of this description and drawing since, while most errors are made by mistaking type-1 for type-5 stools, dispersion is highly relevant for types 2 and 6. Since drawings were used as references in the study, improved illustrations are probably key elements in preventing such misinterpretations. The fact that nurses - allegedly highly experienced in fecal observation - showed such poor concordance is of concern. Obviously, type 7 ("fluid stools with no solid lumps") and type 4 ("smooth, soft, long, sausage-like stools") were best correlated by all three groups. The method to decide which legend corresponded to which drawing (only one definition for each illustration) is most adequate for correctly assessing this scale. As the authors themselves state, their results prompt a redefinition of some types in the scale in order to identify and correct the grounds on which discord arose. Even for a relatively small population sample, the test-retest strategy showed an overall consistency of 84.4% and a kappa index of 0.816, which demonstrates the scale's excellent stability. The fact that worst results are obtained in older patients with lower education levels is also of concern if one considers that this scale will be usually employed with no help from healthcare staff. The reasons for the above deficiencies should possibly be highlighted; whether the issue lies with the legends, the illustrations, or both must be identified, and whether the selected population can be extrapolated to the general population should be pondered. A major bias in this study is that two thirds of the study population is healthcare professionals, theoretically experienced in the terms, inspection, and data collection regarding stool form. In assessing results from healthcare professional knowledge on their experience in areas such as pediatrics, gastroenterology, internal medicine, or geriatrics, where observation and recording of stool characteristics is highly common, should prove essential. This would explain the fact that the results obtained among physicians were substantially better than those recorded from nurses and patients. The above comments on discordant results and the listed biases, also identified by the authors, should prompt further studies to resolve these concerns; these will no doubt shed light on the current issues, and studies in patients or the general population will no doubt be safely performed using the Bristol scale.

References

1. Davies GJ, Crowder M, Reid B, Dickerson JW. Bowel function measurements of individuals with different eating patterns. Gut 1986; 27: 164-9.        [ Links ]

2. Aichbichler BW, Wenzl HH, Santa Ana CA, Porter JL, Schiller LR, Fordtran JS. A comparison of stool characteristics from normal and constipated people. Dig Dis Sci 1998; 43: 2353-62.        [ Links ]

3. O'Donnell LJD, Heaton KW. Pseudo-diarrhea in the irritable bowel syndrome: patients´ records of stool form reflect transit time while stool frequency does not. Gut 1988; 29: A1455.        [ Links ]

4. Heaton KW, Radvan J, Cripps H, Mountford RA, Braddon FE, Hughes AO. Defecation frequency and timing, and stool form in the general population: a prospective study. Gut 1992; 33: 818-24.        [ Links ]

5. Probert CJS, Emmett PM, Heaton KW. Intestinal transit time in the population calculated from self made observations of defecation. J Epidemiol Community Health 1993; 47: 331-3.        [ Links ]

6. Heaton KW, O'Donnell LJ. An office guide to whole-gut transit time. Patients' recollection of their stool form. J Clin Gastroenterol 1994; 19: 28-30.        [ Links ]

7. Degen LP, Phillips SF. How well does stool form reflect colonic transit? Gut 1996; 39: 109-13.        [ Links ]

8. Lewis SJ, Heaton KW. Stool form scale as a useful guide to intestinal transit time. Scand J Gastroenterol 1997; 32: 920-4.        [ Links ]

9. Choung RS, Locke GR 3rd, Zinsmeister AR, Schleck CD, Talley NJ. Epidemiology of slow and fast colonic transit using a scale of stool form in a community. Aliment Pharmacol Ther 2007; 26: 1043-50.        [ Links ]

10. Yilmaz S, Dursun M, Ertem M, Canoruc F, Turhanoglu A. The epidemiological aspects of irritable bowel syndrome in Southeastern Anatolia: a stratified randomised community-based study. Int J Clin Pract 2005; 59: 361-9.        [ Links ]

11. Adibi P, Behzad E, Pirzadeh S, Mohseni M. Bowel habit reference values and abnormalities in young Iranian healthy adults. Dig Dis Sci 2007; 52: 1810-3.        [ Links ]

12. Ersryd A, Posserud I, Abrahamsson H, Simrém M. Subtyping the irritable bowel syndrome by predominant bowel habit: Rome II versus Rome III. Aliment Pharmacol Ther 2007; 26: 953-61.        [ Links ]

13. Tinmouth J, Tomlinson G, Kandel G, Walmsley S, Steinhart H, Glazier R. Evaluation of stool frequency and stool form as measures of HIV-related diarrhea. HIV Clin Trials 2007; 8: 421-8.        [ Links ]

14. Macmillan AK, Merrie AE, Marshall RJ, Parry BR. Design and validation of a comprehensive fecal incontinence questionnaire. Dis Colon Rectum 2008; 51: 1502-22.        [ Links ]

15. Longstreth GF, Thompson WG, Chey WD, Houghton LA, Mearin F, Spiller RC. Functional bowel disorders. Gastroenterology 2006; 130: 1480-91.        [ Links ]

16. Parés D, Comas M, Dorcaratto D, Araujo MI, Vial M, Bohle B, et al. Adaptation and validation of the Bristol scale form into Spanish language among health professionals and patients. Rev Esp Enferm Dig 2009; 101(5): 312-6.        [ Links ]