Inflammatory bowel disease (IBD) is the umbrella term used to describe Crohns disease (CD), ulcerative colitis (UC), and indeterminate colitis (IC). All of which have both significant similarities and differences in their aetiology, presentation, diagnosis and management.1Faecal inflammatory markers are used to assist diagnosis and management of IBD. Of these, calprotectin is the most useful with a wide range of quoted sensitivities and specificities. One meta-analysis (13 studies, 1,041 patients) published in the British Medical Journal in 2010, estimated the sensitivity to be 93%, and specificity 96% for detecting inflammation.2 Von Roon et alfound similar results of 95% and 91%, respectively, in their meta-analysis (30 studies, 5,983 patients) published in the American Journal of Gastroenterology in 2007.3 Specificity of faecal calprotectin for detecting IBD is much lower - due to the various other causes of an elevated level, eg infection, non-steroidal anti-inflammatories, colorectal cancer.4 Despite this, due to its high sensitivity, faecal calprotectin testing presents itself as a highly useful investigation especially in the primary care setting.Calprotectin is found in the cytosol of inflammatory cells and acts as a bacteriostatic cytokine-like substance.4,5 Only a small amount (5g minimum) of stool is required for the test and it is extremely stable at room temperature, so is able to be stored for up to 7 days without diminishing its accuracy.4,5 This enables a lag time between collection and processing-extremely useful in the primary care setting, and similarly so in New Zealand, enabling samples to be processed at non-local labs eg, currently, Canterbury Health Laboratories in Christchurch contracts to Taranaki DHB. It is relatively expensive ($90 per assay) and so is not without some financial implications.The question of the utility of faecal calprotectin has been raised ever since it started to be used more widely, along with the question surrounding the current accepted reference range and if this is appropriate or not. A number of audits, primarily based in the National Health System (NHS) in the UK, have attempted to assess this. A similar audit to this, performed by Allison et al at the Royal Gwent Hospital in Newport, UK, was published in Gut in 2013.6 The authors examined faecal calprotectin samples for 266 patients over a two-year period (Feb 2010-April 2012), who were previously not known to have IBD. Of these, 155 had a normal result; however, management was deemed unaltered in 126 (81%) of the patients, of whom 50 went on to have a (normal) colonoscopy.6 Of the 29 patients where the test was thought to alter management, only 17 were spared the need for endoscopic evaluation. For those 98 patients with an elevated result, 60 (61%) of the patients had unaltered management due to various reasons (eg, further investigation indicated anyway by way of symptoms/bloods), and 27 (28%) were falsely elevated, demonstrated by subsequent negative endoscopic evaluation.6 The same group did another audit looking at faecal calprotectin testing on patients with known IBD.7 Over the same time period, 98 patients were identified with 71% (n=70) having an elevated result. Of these, only nine (13%) had a change in their management.7In 2013, Seenan et al attempted to determine the diagnostic yield of further investigation in patients with lower GI symptoms and mildly elevated faecal calprotectin.8 163 patients (most of who presented with diarrhoea or abdominal pain) were included, after having had a faecal calprotectin mildly positive (between 100-200mcg/g). Of these, 131 went on to have colonoscopy, and only 23 (18%) demonstrated abnormalities.8 Interestingly, IBD was the final diagnosis in only 1.8% (n=3) of the patients investigated, with almost half (48.5%) diagnosed with IBS.8 The negative predictive value of faecal calprotectin from 100-200mcg/g for excluding a significant GI pathology was high, at 98% (for IBD, adenoma, or colorectal carcinoma), casting doubt on the standard laboratory reference range (normal range <50mcg/g).8There have been no such audits done on the use of faecal calprotectin testing in the Taranaki DHB. The aim of this audit was to determine the current rate of request of faecal calprotectin testing within the Taranaki DHB, to determine the clinical indications for requesting this test, and the outcomes for the patients involved. It will provide a baseline set of data with provision of recommendations to improve the service, and a plan to re-audit after recommendations are put in place.MethodPatients were identified as having had a faecal calprotectin test between 1 July 2013 and 31 December 2013 from records kept by Taranaki Medlab, based in New Plymouth (the receiving centre for all Taranaki DHB faecal calprotectin tests). A search of their electronic clinical records was performed to identify the reason for which they received the test. This was achieved by examining a number of potential resources, such as electronic discharge summaries, clinic letters, scanned clinical notes, and indications stated on investigation reports, such as CT abdomens and endoscopies. Information was also extracted on other investigations done at the time of the test, such as stool samples and imaging. Follow-up documentation from the same resources was also examined. Further information was sought from general practitioners, who were approached and requested to provide the relevant clinical documentation as to the events surrounding and precipitating the request of faecal calprotectin testing.ResultsTable 1: gender and age demographics Female 130 Male 76 Age range 17-89 years old - Mean age 53 - Median age 45 - Mode age 50 Figure 1: ordering clinician Figure 2: levels of faecal calprotectin A total of 206 patients over the age of 16 had a faecal calprotectin test during the 6 month period between 1 July 2013 and 31 December 2013. Table 1 and Figure 1 show the basic demographic features and ordering clinician.The range of the level of faecal calprotectin was from 1 to>500mcg/g, with the mean faecal calprotectin 68mcg/g, and median 29mcg/g.The majority of patients had negative results (0-50mcg/g, n=125), or only mildly elevated results (51-100mcg/g, n=23; 0-100mcg/g, n=148). Of those over 100mcg/g, 30 were 100-200mcg/g, and 12 were strongly positive (>500mcg/g) (see Figure 2).A total of 11 patients (5%) tested, had a previous known history of inflammatory bowel disease. Five had a history of bowel cancer.Of the 206 patients tested, 78 (38%) went on to have further investigation, 76 with either upper GI endoscopy (11), colonoscopy (44), sigmoidoscopy (3), or both upper GI endoscopy and colonoscopy (18). Two patients had CT colonographies. Of the 11 patients with a known history of IBD, only one went on to have a colonoscopy (91% avoided invasive investigation), compared to three of the five with a history of bowel cancer progressing for colonoscopy. Six of the patients with known IBD had levels less than 100mcg/g, four had levels from 100-200mcg/g, and one had a level of>500mcg/g this was the sole patient who went on to have a colonoscopy.56 (74%) of the scopes were negative, with 10 identifying diverticulosis, six showing ulcerative colitis, two showing indeterminate colitis, one with Crohns disease, and one diagnosing colorectal cancer after histological examination. Of those scopes showing IBD, all were new diagnoses apart from one, giving an incidence rate of approximately 8% per year significantly higher than the background incidence rate of IBD in the general population (approximately 0.02% per year), but clearly a much more highly selected for patient group.The results were then compared to the original indication given for requesting faecal calprotectin. This was done in an attempt to identify the number of patients where the test was useful, and helped to avoid further unnecessary investigation or intervention, along with those patients where it did not offer any clear clinical benefit due to a number of factors (such as significant symptoms/clinical factors) that would necessitate further investigation irrespective of the result of the faecal calprotectin test. See Figure 4 for a visual representation of this.Figure 3: results of endoscopic evaluation Figure 4: Utility of faecal calprotectin From the 206 patients, 75 were excluded due to lack of information. This left 131 eligible patients. 29 (22%) had faecal calprotectin measured as a method to rule out IBD, successfully avoiding further unnecessary investigation or intervention. For 39 of the patients (30%), the elevated faecal calprotectin prompted onwards referral whether it be surgical or gastroenterology outpatients, or directly to endoscopy. A total of 38 patients (29%) had faecal calprotectin measured without it clearly adding any value to their ongoing management, as the diagnosis was clear (eg, infectious diarrhoea), or their clinical presentation dictated further investigation irrespective of the faecal calprotectin result (most commonly, these patients had both faecal calprotectin and endoscopy/colonoscopy requests made at the same visit). The remaining 19 patients (15%) had endoscopic evaluation following their faecal calprotectin test, but it was unable to clearly define whether or not the test had provided clinical benefit in their case, but all of the endoscopies/colonoscopies were normal in these patients. 11 of these patients had faecal calprotectin levels less than 100mcg/g, and so the assumption may be made that their management was not affected by their essentially normal faecal calprotectin level.DiscussionFaecal calprotectin has been an extremely useful diagnostic and surveillance tool for inflammatory bowel disease for some time now. As with many investigations, it has a number of disadvantages and should always be interpreted within the appropriate clinical situations.The major findings of this audit include a large proportion of patients (29%) not clearly benefitting from the measurement of faecal calprotectin. In a further 11 patients, this was also likely to be the case, making a total of 37% of the examined population a total potential cost of approximately $4,410. This is likely to be an underestimation due to the significant paucity of data that exists. Estimation of the number of patients who underwent subsequent invasive investigation by way of endoscopic bowel evaluation unnecessarily is somewhat more difficult, because when faced with concerning symptoms and an abnormal faecal calprotectin test, the clinician is somewhat obliged to investigate further.This audit does highlight the utility of faecal calprotectin testing when done in the appropriate clinical setting, with a similar proportion (22%) having IBD excluded with a negative test and avoiding more invasive investigation. The audit also demonstrates the potential role for monitoring in patients with known IBD to minimise repeat invasive investigation (91% avoided investigation).The level of faecal calprotectin also appears to have a significant effect on further investigation, along with the rate of diagnosis. This audit demonstrates higher rates of endoscopic evaluation for patients with higher levels or faecal calprotectin, correlating with a higher rate of identification of IBD, especially in those patients with extremely high levels where 88% of the patients were later identified endoscopically to have IBD after an initial faecal calprotectin of>500mcg/g. A very large proportion (86%, n=178) of the patients had levels less than 200. There was a lower rate of investigation in these patients (33%), but also no cases of IBD identified. This supports the observation that a higher level suggests a higher chance of IBD, and also supports the suggestion that a higher cut-off (than the current normal range of 0-50mcg/g) would not compromise the sensitivity and negative predictive value of the test. It most definitely highlights the need for clinical correlation and clinical judgement when interpreting the test.Of note, eight new cases of IBD were identified over the study period of 6 months (an incidence rate of approximately 8% per year). This is likely to be an overestimate of the background incidence rate in Taranaki, due to the highly selected for population included in this audit.There are a number of limitations involved with this study. The study design, being retrospective and observational, means that a number of biasing and confounding features are not accounted for. The information that was able to be gained was done so by examination of patient clinical notes and online results, and so the information extracted from this is under the influence of investigator bias as to the interpretation of various aspects. Along with this, information sets for all patients involved may not be complete. Doing a similar audit, but prospectively, would enable more complete and more efficient information gathering, and would likely reveal more insight into the underlying reasons behind ordering the test. Adding to this, a major limiting factor has been the large percentage (36%, n=75 patients) of patients for whom there was insufficient clinical data available to adequately assess the initial indication, and the subsequent investigations/outcomes for the patient. Unfortunately, this was due to a poor uptake/response rate from practitioners involved. This was despite approval from the local ethics advisor and formal approval from the Health and Disability Ethical Committee. Reasons cited were mainly surrounding patient confidentiality and consent something not usually an issue with clinical audits. Despite this, the data obtained do identify a significant portion of unnecessary tests, and likely provide a true representation of the tests use in current medical practice.Conclusion and recommendationsFaecal calprotectin testing is becoming more common in the assessment of a wide range of gastrointestinal symptoms, and is a very sensitive test when used in the appropriate clinical setting. It also provides a very useful and non-invasive method of monitoring patients with known IBD and, for both of these reasons, can lead to the reduction in the number of potentially avoidable and dangerous invasive investigations.However, it is not specific for IBD and so interpretation of the results must be made within the clinical presentation and alongside other clinical features and investigations as appropriate. More education is needed for those clinicians using the test most frequently particularly in primary care where it will likely prove to have the most benefit.In addition to further education, it would be beneficial to engage primary care providers and work together to develop a locally agreed-upon policy framework for requesting faecal calprotectin with clear recommendations on indications and instances when the test would be beneficial, along with clear recommendations on further intervention and management following either a normal or abnormal test. Once established, this could be easily implemented in Taranaki due to the single receiving centre (as mentioned above, Taranaki Medlab). Repeat audit following this is recommended to assess whether requests for faecal calprotectin are becoming more appropriate.

Inflammatory bowel disease (IBD) is the umbrella term used to describe Crohns disease (CD), ulcerative colitis (UC), and indeterminate colitis (IC). All of which have both significant similarities and differences in their aetiology, presentation, diagnosis and management.1Faecal inflammatory markers are used to assist diagnosis and management of IBD. Of these, calprotectin is the most useful with a wide range of quoted sensitivities and specificities. One meta-analysis (13 studies, 1,041 patients) published in the British Medical Journal in 2010, estimated the sensitivity to be 93%, and specificity 96% for detecting inflammation.2 Von Roon et alfound similar results of 95% and 91%, respectively, in their meta-analysis (30 studies, 5,983 patients) published in the American Journal of Gastroenterology in 2007.3 Specificity of faecal calprotectin for detecting IBD is much lower - due to the various other causes of an elevated level, eg infection, non-steroidal anti-inflammatories, colorectal cancer.4 Despite this, due to its high sensitivity, faecal calprotectin testing presents itself as a highly useful investigation especially in the primary care setting.Calprotectin is found in the cytosol of inflammatory cells and acts as a bacteriostatic cytokine-like substance.4,5 Only a small amount (5g minimum) of stool is required for the test and it is extremely stable at room temperature, so is able to be stored for up to 7 days without diminishing its accuracy.4,5 This enables a lag time between collection and processing-extremely useful in the primary care setting, and similarly so in New Zealand, enabling samples to be processed at non-local labs eg, currently, Canterbury Health Laboratories in Christchurch contracts to Taranaki DHB. It is relatively expensive ($90 per assay) and so is not without some financial implications.The question of the utility of faecal calprotectin has been raised ever since it started to be used more widely, along with the question surrounding the current accepted reference range and if this is appropriate or not. A number of audits, primarily based in the National Health System (NHS) in the UK, have attempted to assess this. A similar audit to this, performed by Allison et al at the Royal Gwent Hospital in Newport, UK, was published in Gut in 2013.6 The authors examined faecal calprotectin samples for 266 patients over a two-year period (Feb 2010-April 2012), who were previously not known to have IBD. Of these, 155 had a normal result; however, management was deemed unaltered in 126 (81%) of the patients, of whom 50 went on to have a (normal) colonoscopy.6 Of the 29 patients where the test was thought to alter management, only 17 were spared the need for endoscopic evaluation. For those 98 patients with an elevated result, 60 (61%) of the patients had unaltered management due to various reasons (eg, further investigation indicated anyway by way of symptoms/bloods), and 27 (28%) were falsely elevated, demonstrated by subsequent negative endoscopic evaluation.6 The same group did another audit looking at faecal calprotectin testing on patients with known IBD.7 Over the same time period, 98 patients were identified with 71% (n=70) having an elevated result. Of these, only nine (13%) had a change in their management.7In 2013, Seenan et al attempted to determine the diagnostic yield of further investigation in patients with lower GI symptoms and mildly elevated faecal calprotectin.8 163 patients (most of who presented with diarrhoea or abdominal pain) were included, after having had a faecal calprotectin mildly positive (between 100-200mcg/g). Of these, 131 went on to have colonoscopy, and only 23 (18%) demonstrated abnormalities.8 Interestingly, IBD was the final diagnosis in only 1.8% (n=3) of the patients investigated, with almost half (48.5%) diagnosed with IBS.8 The negative predictive value of faecal calprotectin from 100-200mcg/g for excluding a significant GI pathology was high, at 98% (for IBD, adenoma, or colorectal carcinoma), casting doubt on the standard laboratory reference range (normal range <50mcg/g).8There have been no such audits done on the use of faecal calprotectin testing in the Taranaki DHB. The aim of this audit was to determine the current rate of request of faecal calprotectin testing within the Taranaki DHB, to determine the clinical indications for requesting this test, and the outcomes for the patients involved. It will provide a baseline set of data with provision of recommendations to improve the service, and a plan to re-audit after recommendations are put in place.MethodPatients were identified as having had a faecal calprotectin test between 1 July 2013 and 31 December 2013 from records kept by Taranaki Medlab, based in New Plymouth (the receiving centre for all Taranaki DHB faecal calprotectin tests). A search of their electronic clinical records was performed to identify the reason for which they received the test. This was achieved by examining a number of potential resources, such as electronic discharge summaries, clinic letters, scanned clinical notes, and indications stated on investigation reports, such as CT abdomens and endoscopies. Information was also extracted on other investigations done at the time of the test, such as stool samples and imaging. Follow-up documentation from the same resources was also examined. Further information was sought from general practitioners, who were approached and requested to provide the relevant clinical documentation as to the events surrounding and precipitating the request of faecal calprotectin testing.ResultsTable 1: gender and age demographics Female 130 Male 76 Age range 17-89 years old - Mean age 53 - Median age 45 - Mode age 50 Figure 1: ordering clinician Figure 2: levels of faecal calprotectin A total of 206 patients over the age of 16 had a faecal calprotectin test during the 6 month period between 1 July 2013 and 31 December 2013. Table 1 and Figure 1 show the basic demographic features and ordering clinician.The range of the level of faecal calprotectin was from 1 to>500mcg/g, with the mean faecal calprotectin 68mcg/g, and median 29mcg/g.The majority of patients had negative results (0-50mcg/g, n=125), or only mildly elevated results (51-100mcg/g, n=23; 0-100mcg/g, n=148). Of those over 100mcg/g, 30 were 100-200mcg/g, and 12 were strongly positive (>500mcg/g) (see Figure 2).A total of 11 patients (5%) tested, had a previous known history of inflammatory bowel disease. Five had a history of bowel cancer.Of the 206 patients tested, 78 (38%) went on to have further investigation, 76 with either upper GI endoscopy (11), colonoscopy (44), sigmoidoscopy (3), or both upper GI endoscopy and colonoscopy (18). Two patients had CT colonographies. Of the 11 patients with a known history of IBD, only one went on to have a colonoscopy (91% avoided invasive investigation), compared to three of the five with a history of bowel cancer progressing for colonoscopy. Six of the patients with known IBD had levels less than 100mcg/g, four had levels from 100-200mcg/g, and one had a level of>500mcg/g this was the sole patient who went on to have a colonoscopy.56 (74%) of the scopes were negative, with 10 identifying diverticulosis, six showing ulcerative colitis, two showing indeterminate colitis, one with Crohns disease, and one diagnosing colorectal cancer after histological examination. Of those scopes showing IBD, all were new diagnoses apart from one, giving an incidence rate of approximately 8% per year significantly higher than the background incidence rate of IBD in the general population (approximately 0.02% per year), but clearly a much more highly selected for patient group.The results were then compared to the original indication given for requesting faecal calprotectin. This was done in an attempt to identify the number of patients where the test was useful, and helped to avoid further unnecessary investigation or intervention, along with those patients where it did not offer any clear clinical benefit due to a number of factors (such as significant symptoms/clinical factors) that would necessitate further investigation irrespective of the result of the faecal calprotectin test. See Figure 4 for a visual representation of this.Figure 3: results of endoscopic evaluation Figure 4: Utility of faecal calprotectin From the 206 patients, 75 were excluded due to lack of information. This left 131 eligible patients. 29 (22%) had faecal calprotectin measured as a method to rule out IBD, successfully avoiding further unnecessary investigation or intervention. For 39 of the patients (30%), the elevated faecal calprotectin prompted onwards referral whether it be surgical or gastroenterology outpatients, or directly to endoscopy. A total of 38 patients (29%) had faecal calprotectin measured without it clearly adding any value to their ongoing management, as the diagnosis was clear (eg, infectious diarrhoea), or their clinical presentation dictated further investigation irrespective of the faecal calprotectin result (most commonly, these patients had both faecal calprotectin and endoscopy/colonoscopy requests made at the same visit). The remaining 19 patients (15%) had endoscopic evaluation following their faecal calprotectin test, but it was unable to clearly define whether or not the test had provided clinical benefit in their case, but all of the endoscopies/colonoscopies were normal in these patients. 11 of these patients had faecal calprotectin levels less than 100mcg/g, and so the assumption may be made that their management was not affected by their essentially normal faecal calprotectin level.DiscussionFaecal calprotectin has been an extremely useful diagnostic and surveillance tool for inflammatory bowel disease for some time now. As with many investigations, it has a number of disadvantages and should always be interpreted within the appropriate clinical situations.The major findings of this audit include a large proportion of patients (29%) not clearly benefitting from the measurement of faecal calprotectin. In a further 11 patients, this was also likely to be the case, making a total of 37% of the examined population a total potential cost of approximately $4,410. This is likely to be an underestimation due to the significant paucity of data that exists. Estimation of the number of patients who underwent subsequent invasive investigation by way of endoscopic bowel evaluation unnecessarily is somewhat more difficult, because when faced with concerning symptoms and an abnormal faecal calprotectin test, the clinician is somewhat obliged to investigate further.This audit does highlight the utility of faecal calprotectin testing when done in the appropriate clinical setting, with a similar proportion (22%) having IBD excluded with a negative test and avoiding more invasive investigation. The audit also demonstrates the potential role for monitoring in patients with known IBD to minimise repeat invasive investigation (91% avoided investigation).The level of faecal calprotectin also appears to have a significant effect on further investigation, along with the rate of diagnosis. This audit demonstrates higher rates of endoscopic evaluation for patients with higher levels or faecal calprotectin, correlating with a higher rate of identification of IBD, especially in those patients with extremely high levels where 88% of the patients were later identified endoscopically to have IBD after an initial faecal calprotectin of>500mcg/g. A very large proportion (86%, n=178) of the patients had levels less than 200. There was a lower rate of investigation in these patients (33%), but also no cases of IBD identified. This supports the observation that a higher level suggests a higher chance of IBD, and also supports the suggestion that a higher cut-off (than the current normal range of 0-50mcg/g) would not compromise the sensitivity and negative predictive value of the test. It most definitely highlights the need for clinical correlation and clinical judgement when interpreting the test.Of note, eight new cases of IBD were identified over the study period of 6 months (an incidence rate of approximately 8% per year). This is likely to be an overestimate of the background incidence rate in Taranaki, due to the highly selected for population included in this audit.There are a number of limitations involved with this study. The study design, being retrospective and observational, means that a number of biasing and confounding features are not accounted for. The information that was able to be gained was done so by examination of patient clinical notes and online results, and so the information extracted from this is under the influence of investigator bias as to the interpretation of various aspects. Along with this, information sets for all patients involved may not be complete. Doing a similar audit, but prospectively, would enable more complete and more efficient information gathering, and would likely reveal more insight into the underlying reasons behind ordering the test. Adding to this, a major limiting factor has been the large percentage (36%, n=75 patients) of patients for whom there was insufficient clinical data available to adequately assess the initial indication, and the subsequent investigations/outcomes for the patient. Unfortunately, this was due to a poor uptake/response rate from practitioners involved. This was despite approval from the local ethics advisor and formal approval from the Health and Disability Ethical Committee. Reasons cited were mainly surrounding patient confidentiality and consent something not usually an issue with clinical audits. Despite this, the data obtained do identify a significant portion of unnecessary tests, and likely provide a true representation of the tests use in current medical practice.Conclusion and recommendationsFaecal calprotectin testing is becoming more common in the assessment of a wide range of gastrointestinal symptoms, and is a very sensitive test when used in the appropriate clinical setting. It also provides a very useful and non-invasive method of monitoring patients with known IBD and, for both of these reasons, can lead to the reduction in the number of potentially avoidable and dangerous invasive investigations.However, it is not specific for IBD and so interpretation of the results must be made within the clinical presentation and alongside other clinical features and investigations as appropriate. More education is needed for those clinicians using the test most frequently particularly in primary care where it will likely prove to have the most benefit.In addition to further education, it would be beneficial to engage primary care providers and work together to develop a locally agreed-upon policy framework for requesting faecal calprotectin with clear recommendations on indications and instances when the test would be beneficial, along with clear recommendations on further intervention and management following either a normal or abnormal test. Once established, this could be easily implemented in Taranaki due to the single receiving centre (as mentioned above, Taranaki Medlab). Repeat audit following this is recommended to assess whether requests for faecal calprotectin are becoming more appropriate.