The Journal of Family Practice

CASE › It was a clinical conundrum. A 2011 case study¹ described a 33-year-old woman with a 10-year history of progressive, debilitating pain and weakness. The patient had not received a unifying diagnosis or effective treatment despite multiple diagnostic tests and different recommendations from multiple specialists. The diagnosis remained elusive until a rheumatologist agreed to reexamine the case.

While reviewing the woman’s thick chart, the rheumatologist noted a series of negative results from upper and lower endoscopies and abdominal scans. Further investigation revealed an almost obfuscated clue—blood tests performed 2 years earlier that were positive for celiac disease (CD). However, a small intestine biopsy, which normally is done to confirm the diagnosis, was never performed.

The rheumatologist made a tentative diagnosis of CD and referred her to a nutritionist, who recommended the patient adhere to a strict gluten-free diet. Within 3 months, the patient experienced marked improvement and returned to work.

CD is an often-missed diagnosis. According to a study based on National Health and Nutrition Examination Survey data, only 17% of patients with CD are aware they have the disease.² As such, it is imperative that primary care physicians familiarize themselves with CD’s myriad clinical presentations, diagnosis, and treatment.^3-6

Gluten triggers an immune response
 in genetically susceptible patients


CD initially was known as “celiac sprue” because it shares characteristics with tropical sprue—diarrhea, malabsorption, and emaciation. It is a unique T-cell autoimmune enteropathy that is precipitated in genetically susceptible individuals by the ingestion of gluten, the major storage protein of wheat, barley, and rye.^3,7

Upon ingestion, gluten breaks down to gliadin, which provokes an immune response in the intestinal mucosa of patients with CD. This response results in an inflammatory reaction, primarily in the upper small intestine, that destroys the absorption surface and causes villous atrophy, leading to nutrient malabsorption and chronic diarrhea.⁸ CD is associated with significant morbidity due to an abnormal excretion of fat (steatorrhea) and varying degrees of malabsorption of vitamins A, D, and K, as well as B complex vitamins including B12 and folate; carbohydrates; protein; water; and minerals such as magnesium, calcium, and iron.⁹

CD develops only in individuals who possess alleles that encode for HLA-DQ2 or HLA-DQ8 proteins, products of 2 of the HLA genes. And while 30% of Caucasians carry the HLA-DQ2 allele and virtually 100% consume wheat, only 1 in 100 will develop CD.^3,10,11 Although the genes are necessary, it is the interplay between genes (both HLA and non-HLA associated) and environment (ie, gluten) that leads to the intestinal mucosa damage typical of the disease. The HLA-DQ region also is associated with increased risk of type 1 diabetes, which might explain the correlation of CD to a host of other autoimmune disorders, including Graves’ disease and rheumatoid arthritis.^8,10,11

Increased prevalence reflects 
better recognition of celiac disease

CD affects .6% to 1% of the population worldwide, with wide regional variation.3 Before the development of serologic assays in the 1970s, CD was a clinical diagnosis based on classic symptoms. With the advent of assays for immunoglobulin A (IgA) antibodies, the prevalence of CD has drastically increased to the current estimates of 1:250 to 1:500.^4,5 The prevalence will continue to increase as clinicians become more aware of the different presentations of the disease, which are described below.

CD runs in families. Most patients with CD have a family history of the disease based on inheritance of the HLA alleles. A US study determined that the prevalence of CD was 1:22 in first-degree relatives and 1:39 in second-degree relatives of patients with biopsy-proven CD.¹²

Less than half of patients 
have GI symptoms


The classic presentation of CD involves a constellation of signs and symptoms of malabsorption: diarrhea, muscle wasting, and weight loss. Other typical gastrointestinal (GI) symptoms include bloating, flatulence, and abdominal pain.

It is the interplay between genes and the environment that leads to the intestinal mucosa damage typical of the disease. Recognizing CD can be challenging, however, because <50% of patients diagnosed with CD present with these classic GI symptoms.³ About 50% of CD patients present with extra-intestinal symptoms, such as iron deficiency anemia, aphthous stomatitis, chronic fatigue, osteopenia, and dental enamel hypoplasia.^3,8,13 Other possible non-GI symptoms include abnormal liver function test results and skin disorders such as dermatitis herpetiformis, a pruritic rash with cutaneous IgA deposits.^3,8 In addition, many patients are asymptomatic.¹⁴ This highly variable clinical picture is due to the genetic and immunologic basis of the disease, extent of mucosal injury, and patients’ dietary habits, gender, and age of onset.¹⁵ A common clue that suggests a patient may have CD is unexplained iron deficiency anemia that does not improve with oral iron supplementation.^4,13

For patients 
who are consuming a diet that includes gluten and have symptoms that suggest CD, the ACG guidelines recommend initial testing for IgA tissue transglutaminase (tTG) antibodies. Because symptoms may be intermittent, a patient may delay seeking care until he or she develops secondary manifestations, which often are debilitating and overshadow the GI complaints. Chronic complications of untreated CD include lymphoma and adenocarcinomas of the jejunum, recurrent miscarriages, neurologic disorders, osteoporosis, and hyposplenism.^3,4,8

Since CD can manifest with widely varying symptoms, some researchers believe the disease should be classified into 3 categories based on presentation: classic CD, which presents with diarrhea, weight loss, malabsorption, and vitamin deficiency; atypical CD, which presents with minimal GI symptoms but can include anemia, neurologic symptoms, arthritis, or infertility; and asymptomatic CD, which typically displays no symptoms but usually is identified on incidental screening.^3,8,16 Non-celiac gluten sensitivity is a distinct condition in which the body reacts adversely to gluten; it is not an autoimmune disease with an inflammatory response.

Order serologic testing 
for at-risk patients


Because CD remains underdiagnosed,¹⁶ taking a thorough family history and dietary history and making sure to at least consider CD as a part of a differential diagnosis is important. Although population-based screening has been proposed, its benefits and cost-effectiveness remain unproven. As a result, serologic testing of at-risk groups—individuals with conditions known to be associated with CD—remains the current standard.³ The TABLE lists groups for whom serologic testing for CD is indicated.^16,17

In addition, the American College of Gastroenterology (ACG) and the North American Society for Pediatric Gastroenterology, Hepatology, and Nutrition (NASPGHAN) provide guidance on the diagnosis and treatment of adults and children with CD. (An ACG diagnostic algorithm is available at http://www.nature.com/ajg/journal/v108/n5/pdf/ajg201379a.pdf.)

Adults. For patients who are consuming a diet that includes gluten and have symptoms that suggest CD, the ACG guidelines recommend initial testing for IgA tissue transglutaminase (tTG) antibodies.¹⁶ The IgA tTG has a sensitivity and specificity >95%.¹⁶ An alternative test, the IgA endomysial (IgA EMA) test, has similar sensitivity but is time-consuming and its accuracy depends on the experience and skill of the laboratory technician. A negative result for either test has a high negative predictive value for CD.^3,16

IgA deficiency is much more common in patients with CD than in the general population and can result in a false negative test for tTG and EMA. Therefore, consider taking a baseline IgA measurement first. If the patient has an IgA deficiency, the test you’ll use next will change: The preferred test for CD is either immunoglobulin G (IgG) tTG or IgG deamidated gliadin peptides (DGP).^3,16

If a patient is already gluten-free... To rule out CD in patients who are already consuming a gluten-free diet, order HLA-DQ2 and HLA-DQ8 testing because these markers have a specificity >99%; if the HLA test is negative, the disease is excluded.^8,16

Children. NASPGHAN recommends taking a baseline IgA measurement in children at risk for CD and then testing for IgA tTG antibodies, but not until patients are 3 years old and have been on a diet that includes gluten for at least 1 year.¹⁷ Repeat testing at a later date it is recommended for those with negative results because some evidence suggests that in certain patients, later serologic testing will be positive. Alternatively, you may offer HLA testing. If the HLA test is negative, CD can be excluded >99% of the time.

Diagnosis usually is confirmed 
by intestinal biopsy


Positive results on serologic testing should be confirmed with a biopsy of the small bowel; findings characteristic of CD include an increased number of intraepithelial lymphocytes (>25 per 100 enterocytes), elongation of the crypts, and partial to total villous atrophy.⁴ Final confirmation of CD is resolution of symptoms by consuming a gluten-free diet.^3,8

Alternate approaches to confirming the diagnosis. Although intestinal biopsy has long been considered the gold standard for diagnosis of CD, this may change. In 2012, the European Society for Paediatric Gastroenterology, Hepatology, and Nutrition proposed that the biopsy may not be necessary in children with the following 3 characteristics: classic intestinal symptoms of CD, IgA tTG levels >10 times higher than normal, and a positive HLA-DQ2.¹⁸

Catassi and Fasano¹⁹ have proposed shifting from relying on algorithms and intestinal biopsy to a quantitative approach. They suggest using the “4 out of 5” rule, meaning the diagnosis of CD can be confirmed if at least 4 of the following 5 criteria are satisfied: typical CD symptoms, a positive IgA tTG, a positive HLA-DQ2 or -DQ8, celiac enteropathy on small bowel biopsy, and response to a gluten-free diet.¹⁹

The only proven treatment:
 A gluten-free diet

Non-celiac gluten sensitivity is a distinct condition in which the body reacts adversely to gluten; it is not an autoimmune disease with an inflammatory response.Lifelong adherence to a gluten-free diet is the only effective treatment for CD.^14,16 Previously, patients with CD were advised to also avoid oats, but most evidence supports the safety of oats (<2 oz/d), provided there is no cross-contamination with gluten.¹⁴ Adhering to a strict gluten-free diet can be challenging because cereal flours are ubiquitous in western foods, and some foods may be cross-contaminated. The Celiac Disease Foundation (http://www.celiac.org) offers guidance on maintaining a gluten-free diet.

Because avoiding gluten has become popular even among people who don’t have CD, product labeling that includes information on gluten content has become pervasive. However, determining which items contain gluten depends on accurate labeling, a standard that often is not met in many countries; in the United States, such labeling began to be phased in starting in July 2014.²⁰ As a result, CD patients may unwittingly be exposed to gluten over the long term, which can result in greater morbidity and mortality. Unless a food is labeled “gluten-free," it is best to check with the manufacturer.

Compliance with a gluten-free diet can be monitored by following IgA tTG titers every 1 to 2 years, as these values normalize after a patient has been adhering to the diet for 6 to 24 months.^3,16

In addition to lifelong adherence to a gluten-free diet, a National Institutes of Health Consensus Development Conference recommended that management of patients with CD should include²¹:

• consultation with a skilled dietitian

• education about the disease
• 
continuous long-term follow-up by a multidisciplinary team
• 
identification and follow-up of abnormalities found at baseline, such as abnormal liver function test results

• treatment of nutritional deficiencies.

The ACG also recommends that CD patients receive a dual energy x-ray scan for follow-up of osteopenia and a pneumococcal vaccine because functional hyposplenism is associated with CD, and pneumonia is a common complication of hyposplenism.^16,18

Should you recommend a gluten-free diet for other patients? Because avoiding gluten is now popular and many gluten-free products are marketed as “health food,” physicians may be reluctant to recommend a gluten-free diet for patients who have vague abdominal symptoms but negative CD test results. Despite the current popularity of “going gluten-free,” the reality is that in addition to CD, many other diseases may be helped by a gluten-free diet, such as dermatitis herpetiformis, irritable bowel syndrome, and neurologic diseases such as gluten-sensitive ataxia.¹⁹ In the end, whether to adopt a gluten-free diet is a decision that you and your patient will need to make together.

Researchers are searching 
for additional treatments


Because many patients find it difficult to adhere to a gluten-free diet, researchers are investigating several alternative treatments, including a derivative from cholera toxin that inhibits the opening of intestinal epithelial junctions, thereby reducing the resultant inflammatory response, and a desensitizing vaccine.^19,22,23 Another intriguing approach involves using the parasite Necator americanus to modulate the immune response to gluten.18 Finally, certain infant feeding practices, including breastfeeding and delaying introduction of gluten to the diet, may minimize the risk of developing CD. (See "A link between infant feeding practices and the risk of CD?" above.^22-25)

CORRESPONDENCE
Patrick T. Dowling, MD, MPH, Department of Family Medicine, 50-078 Center for Health Sciences, David Geffen School of Medicine at UCLA, Los Angeles, CA 90095-1683; pdowling@mednet.ucla.edu

CASE › It was a clinical conundrum. A 2011 case study¹ described a 33-year-old woman with a 10-year history of progressive, debilitating pain and weakness. The patient had not received a unifying diagnosis or effective treatment despite multiple diagnostic tests and different recommendations from multiple specialists. The diagnosis remained elusive until a rheumatologist agreed to reexamine the case.

While reviewing the woman’s thick chart, the rheumatologist noted a series of negative results from upper and lower endoscopies and abdominal scans. Further investigation revealed an almost obfuscated clue—blood tests performed 2 years earlier that were positive for celiac disease (CD). However, a small intestine biopsy, which normally is done to confirm the diagnosis, was never performed.

The rheumatologist made a tentative diagnosis of CD and referred her to a nutritionist, who recommended the patient adhere to a strict gluten-free diet. Within 3 months, the patient experienced marked improvement and returned to work.

CD is an often-missed diagnosis. According to a study based on National Health and Nutrition Examination Survey data, only 17% of patients with CD are aware they have the disease.² As such, it is imperative that primary care physicians familiarize themselves with CD’s myriad clinical presentations, diagnosis, and treatment.^3-6

Gluten triggers an immune response
 in genetically susceptible patients


CD initially was known as “celiac sprue” because it shares characteristics with tropical sprue—diarrhea, malabsorption, and emaciation. It is a unique T-cell autoimmune enteropathy that is precipitated in genetically susceptible individuals by the ingestion of gluten, the major storage protein of wheat, barley, and rye.^3,7

Upon ingestion, gluten breaks down to gliadin, which provokes an immune response in the intestinal mucosa of patients with CD. This response results in an inflammatory reaction, primarily in the upper small intestine, that destroys the absorption surface and causes villous atrophy, leading to nutrient malabsorption and chronic diarrhea.⁸ CD is associated with significant morbidity due to an abnormal excretion of fat (steatorrhea) and varying degrees of malabsorption of vitamins A, D, and K, as well as B complex vitamins including B12 and folate; carbohydrates; protein; water; and minerals such as magnesium, calcium, and iron.⁹

CD develops only in individuals who possess alleles that encode for HLA-DQ2 or HLA-DQ8 proteins, products of 2 of the HLA genes. And while 30% of Caucasians carry the HLA-DQ2 allele and virtually 100% consume wheat, only 1 in 100 will develop CD.^3,10,11 Although the genes are necessary, it is the interplay between genes (both HLA and non-HLA associated) and environment (ie, gluten) that leads to the intestinal mucosa damage typical of the disease. The HLA-DQ region also is associated with increased risk of type 1 diabetes, which might explain the correlation of CD to a host of other autoimmune disorders, including Graves’ disease and rheumatoid arthritis.^8,10,11

Increased prevalence reflects 
better recognition of celiac disease

CD affects .6% to 1% of the population worldwide, with wide regional variation.3 Before the development of serologic assays in the 1970s, CD was a clinical diagnosis based on classic symptoms. With the advent of assays for immunoglobulin A (IgA) antibodies, the prevalence of CD has drastically increased to the current estimates of 1:250 to 1:500.^4,5 The prevalence will continue to increase as clinicians become more aware of the different presentations of the disease, which are described below.

CD runs in families. Most patients with CD have a family history of the disease based on inheritance of the HLA alleles. A US study determined that the prevalence of CD was 1:22 in first-degree relatives and 1:39 in second-degree relatives of patients with biopsy-proven CD.¹²

Less than half of patients 
have GI symptoms


The classic presentation of CD involves a constellation of signs and symptoms of malabsorption: diarrhea, muscle wasting, and weight loss. Other typical gastrointestinal (GI) symptoms include bloating, flatulence, and abdominal pain.

It is the interplay between genes and the environment that leads to the intestinal mucosa damage typical of the disease. Recognizing CD can be challenging, however, because <50% of patients diagnosed with CD present with these classic GI symptoms.³ About 50% of CD patients present with extra-intestinal symptoms, such as iron deficiency anemia, aphthous stomatitis, chronic fatigue, osteopenia, and dental enamel hypoplasia.^3,8,13 Other possible non-GI symptoms include abnormal liver function test results and skin disorders such as dermatitis herpetiformis, a pruritic rash with cutaneous IgA deposits.^3,8 In addition, many patients are asymptomatic.¹⁴ This highly variable clinical picture is due to the genetic and immunologic basis of the disease, extent of mucosal injury, and patients’ dietary habits, gender, and age of onset.¹⁵ A common clue that suggests a patient may have CD is unexplained iron deficiency anemia that does not improve with oral iron supplementation.^4,13

For patients 
who are consuming a diet that includes gluten and have symptoms that suggest CD, the ACG guidelines recommend initial testing for IgA tissue transglutaminase (tTG) antibodies. Because symptoms may be intermittent, a patient may delay seeking care until he or she develops secondary manifestations, which often are debilitating and overshadow the GI complaints. Chronic complications of untreated CD include lymphoma and adenocarcinomas of the jejunum, recurrent miscarriages, neurologic disorders, osteoporosis, and hyposplenism.^3,4,8

Since CD can manifest with widely varying symptoms, some researchers believe the disease should be classified into 3 categories based on presentation: classic CD, which presents with diarrhea, weight loss, malabsorption, and vitamin deficiency; atypical CD, which presents with minimal GI symptoms but can include anemia, neurologic symptoms, arthritis, or infertility; and asymptomatic CD, which typically displays no symptoms but usually is identified on incidental screening.^3,8,16 Non-celiac gluten sensitivity is a distinct condition in which the body reacts adversely to gluten; it is not an autoimmune disease with an inflammatory response.

Order serologic testing 
for at-risk patients


Because CD remains underdiagnosed,¹⁶ taking a thorough family history and dietary history and making sure to at least consider CD as a part of a differential diagnosis is important. Although population-based screening has been proposed, its benefits and cost-effectiveness remain unproven. As a result, serologic testing of at-risk groups—individuals with conditions known to be associated with CD—remains the current standard.³ The TABLE lists groups for whom serologic testing for CD is indicated.^16,17

In addition, the American College of Gastroenterology (ACG) and the North American Society for Pediatric Gastroenterology, Hepatology, and Nutrition (NASPGHAN) provide guidance on the diagnosis and treatment of adults and children with CD. (An ACG diagnostic algorithm is available at http://www.nature.com/ajg/journal/v108/n5/pdf/ajg201379a.pdf.)

Adults. For patients who are consuming a diet that includes gluten and have symptoms that suggest CD, the ACG guidelines recommend initial testing for IgA tissue transglutaminase (tTG) antibodies.¹⁶ The IgA tTG has a sensitivity and specificity >95%.¹⁶ An alternative test, the IgA endomysial (IgA EMA) test, has similar sensitivity but is time-consuming and its accuracy depends on the experience and skill of the laboratory technician. A negative result for either test has a high negative predictive value for CD.^3,16

IgA deficiency is much more common in patients with CD than in the general population and can result in a false negative test for tTG and EMA. Therefore, consider taking a baseline IgA measurement first. If the patient has an IgA deficiency, the test you’ll use next will change: The preferred test for CD is either immunoglobulin G (IgG) tTG or IgG deamidated gliadin peptides (DGP).^3,16

If a patient is already gluten-free... To rule out CD in patients who are already consuming a gluten-free diet, order HLA-DQ2 and HLA-DQ8 testing because these markers have a specificity >99%; if the HLA test is negative, the disease is excluded.^8,16

Children. NASPGHAN recommends taking a baseline IgA measurement in children at risk for CD and then testing for IgA tTG antibodies, but not until patients are 3 years old and have been on a diet that includes gluten for at least 1 year.¹⁷ Repeat testing at a later date it is recommended for those with negative results because some evidence suggests that in certain patients, later serologic testing will be positive. Alternatively, you may offer HLA testing. If the HLA test is negative, CD can be excluded >99% of the time.

Diagnosis usually is confirmed 
by intestinal biopsy


Positive results on serologic testing should be confirmed with a biopsy of the small bowel; findings characteristic of CD include an increased number of intraepithelial lymphocytes (>25 per 100 enterocytes), elongation of the crypts, and partial to total villous atrophy.⁴ Final confirmation of CD is resolution of symptoms by consuming a gluten-free diet.^3,8

Alternate approaches to confirming the diagnosis. Although intestinal biopsy has long been considered the gold standard for diagnosis of CD, this may change. In 2012, the European Society for Paediatric Gastroenterology, Hepatology, and Nutrition proposed that the biopsy may not be necessary in children with the following 3 characteristics: classic intestinal symptoms of CD, IgA tTG levels >10 times higher than normal, and a positive HLA-DQ2.¹⁸

Catassi and Fasano¹⁹ have proposed shifting from relying on algorithms and intestinal biopsy to a quantitative approach. They suggest using the “4 out of 5” rule, meaning the diagnosis of CD can be confirmed if at least 4 of the following 5 criteria are satisfied: typical CD symptoms, a positive IgA tTG, a positive HLA-DQ2 or -DQ8, celiac enteropathy on small bowel biopsy, and response to a gluten-free diet.¹⁹

The only proven treatment:
 A gluten-free diet

Non-celiac gluten sensitivity is a distinct condition in which the body reacts adversely to gluten; it is not an autoimmune disease with an inflammatory response.Lifelong adherence to a gluten-free diet is the only effective treatment for CD.^14,16 Previously, patients with CD were advised to also avoid oats, but most evidence supports the safety of oats (<2 oz/d), provided there is no cross-contamination with gluten.¹⁴ Adhering to a strict gluten-free diet can be challenging because cereal flours are ubiquitous in western foods, and some foods may be cross-contaminated. The Celiac Disease Foundation (http://www.celiac.org) offers guidance on maintaining a gluten-free diet.

Because avoiding gluten has become popular even among people who don’t have CD, product labeling that includes information on gluten content has become pervasive. However, determining which items contain gluten depends on accurate labeling, a standard that often is not met in many countries; in the United States, such labeling began to be phased in starting in July 2014.²⁰ As a result, CD patients may unwittingly be exposed to gluten over the long term, which can result in greater morbidity and mortality. Unless a food is labeled “gluten-free," it is best to check with the manufacturer.

Compliance with a gluten-free diet can be monitored by following IgA tTG titers every 1 to 2 years, as these values normalize after a patient has been adhering to the diet for 6 to 24 months.^3,16

In addition to lifelong adherence to a gluten-free diet, a National Institutes of Health Consensus Development Conference recommended that management of patients with CD should include²¹:

• consultation with a skilled dietitian

• education about the disease
• 
continuous long-term follow-up by a multidisciplinary team
• 
identification and follow-up of abnormalities found at baseline, such as abnormal liver function test results

• treatment of nutritional deficiencies.

The ACG also recommends that CD patients receive a dual energy x-ray scan for follow-up of osteopenia and a pneumococcal vaccine because functional hyposplenism is associated with CD, and pneumonia is a common complication of hyposplenism.^16,18

Should you recommend a gluten-free diet for other patients? Because avoiding gluten is now popular and many gluten-free products are marketed as “health food,” physicians may be reluctant to recommend a gluten-free diet for patients who have vague abdominal symptoms but negative CD test results. Despite the current popularity of “going gluten-free,” the reality is that in addition to CD, many other diseases may be helped by a gluten-free diet, such as dermatitis herpetiformis, irritable bowel syndrome, and neurologic diseases such as gluten-sensitive ataxia.¹⁹ In the end, whether to adopt a gluten-free diet is a decision that you and your patient will need to make together.

Researchers are searching 
for additional treatments


Because many patients find it difficult to adhere to a gluten-free diet, researchers are investigating several alternative treatments, including a derivative from cholera toxin that inhibits the opening of intestinal epithelial junctions, thereby reducing the resultant inflammatory response, and a desensitizing vaccine.^19,22,23 Another intriguing approach involves using the parasite Necator americanus to modulate the immune response to gluten.18 Finally, certain infant feeding practices, including breastfeeding and delaying introduction of gluten to the diet, may minimize the risk of developing CD. (See "A link between infant feeding practices and the risk of CD?" above.^22-25)

CORRESPONDENCE
Patrick T. Dowling, MD, MPH, Department of Family Medicine, 50-078 Center for Health Sciences, David Geffen School of Medicine at UCLA, Los Angeles, CA 90095-1683; pdowling@mednet.ucla.edu

CASE › It was a clinical conundrum. A 2011 case study¹ described a 33-year-old woman with a 10-year history of progressive, debilitating pain and weakness. The patient had not received a unifying diagnosis or effective treatment despite multiple diagnostic tests and different recommendations from multiple specialists. The diagnosis remained elusive until a rheumatologist agreed to reexamine the case.

While reviewing the woman’s thick chart, the rheumatologist noted a series of negative results from upper and lower endoscopies and abdominal scans. Further investigation revealed an almost obfuscated clue—blood tests performed 2 years earlier that were positive for celiac disease (CD). However, a small intestine biopsy, which normally is done to confirm the diagnosis, was never performed.

The rheumatologist made a tentative diagnosis of CD and referred her to a nutritionist, who recommended the patient adhere to a strict gluten-free diet. Within 3 months, the patient experienced marked improvement and returned to work.

CD is an often-missed diagnosis. According to a study based on National Health and Nutrition Examination Survey data, only 17% of patients with CD are aware they have the disease.² As such, it is imperative that primary care physicians familiarize themselves with CD’s myriad clinical presentations, diagnosis, and treatment.^3-6

Gluten triggers an immune response
 in genetically susceptible patients


CD initially was known as “celiac sprue” because it shares characteristics with tropical sprue—diarrhea, malabsorption, and emaciation. It is a unique T-cell autoimmune enteropathy that is precipitated in genetically susceptible individuals by the ingestion of gluten, the major storage protein of wheat, barley, and rye.^3,7

Upon ingestion, gluten breaks down to gliadin, which provokes an immune response in the intestinal mucosa of patients with CD. This response results in an inflammatory reaction, primarily in the upper small intestine, that destroys the absorption surface and causes villous atrophy, leading to nutrient malabsorption and chronic diarrhea.⁸ CD is associated with significant morbidity due to an abnormal excretion of fat (steatorrhea) and varying degrees of malabsorption of vitamins A, D, and K, as well as B complex vitamins including B12 and folate; carbohydrates; protein; water; and minerals such as magnesium, calcium, and iron.⁹

CD develops only in individuals who possess alleles that encode for HLA-DQ2 or HLA-DQ8 proteins, products of 2 of the HLA genes. And while 30% of Caucasians carry the HLA-DQ2 allele and virtually 100% consume wheat, only 1 in 100 will develop CD.^3,10,11 Although the genes are necessary, it is the interplay between genes (both HLA and non-HLA associated) and environment (ie, gluten) that leads to the intestinal mucosa damage typical of the disease. The HLA-DQ region also is associated with increased risk of type 1 diabetes, which might explain the correlation of CD to a host of other autoimmune disorders, including Graves’ disease and rheumatoid arthritis.^8,10,11

Increased prevalence reflects 
better recognition of celiac disease

CD affects .6% to 1% of the population worldwide, with wide regional variation.3 Before the development of serologic assays in the 1970s, CD was a clinical diagnosis based on classic symptoms. With the advent of assays for immunoglobulin A (IgA) antibodies, the prevalence of CD has drastically increased to the current estimates of 1:250 to 1:500.^4,5 The prevalence will continue to increase as clinicians become more aware of the different presentations of the disease, which are described below.

CD runs in families. Most patients with CD have a family history of the disease based on inheritance of the HLA alleles. A US study determined that the prevalence of CD was 1:22 in first-degree relatives and 1:39 in second-degree relatives of patients with biopsy-proven CD.¹²

Less than half of patients 
have GI symptoms


The classic presentation of CD involves a constellation of signs and symptoms of malabsorption: diarrhea, muscle wasting, and weight loss. Other typical gastrointestinal (GI) symptoms include bloating, flatulence, and abdominal pain.

It is the interplay between genes and the environment that leads to the intestinal mucosa damage typical of the disease. Recognizing CD can be challenging, however, because <50% of patients diagnosed with CD present with these classic GI symptoms.³ About 50% of CD patients present with extra-intestinal symptoms, such as iron deficiency anemia, aphthous stomatitis, chronic fatigue, osteopenia, and dental enamel hypoplasia.^3,8,13 Other possible non-GI symptoms include abnormal liver function test results and skin disorders such as dermatitis herpetiformis, a pruritic rash with cutaneous IgA deposits.^3,8 In addition, many patients are asymptomatic.¹⁴ This highly variable clinical picture is due to the genetic and immunologic basis of the disease, extent of mucosal injury, and patients’ dietary habits, gender, and age of onset.¹⁵ A common clue that suggests a patient may have CD is unexplained iron deficiency anemia that does not improve with oral iron supplementation.^4,13

For patients 
who are consuming a diet that includes gluten and have symptoms that suggest CD, the ACG guidelines recommend initial testing for IgA tissue transglutaminase (tTG) antibodies. Because symptoms may be intermittent, a patient may delay seeking care until he or she develops secondary manifestations, which often are debilitating and overshadow the GI complaints. Chronic complications of untreated CD include lymphoma and adenocarcinomas of the jejunum, recurrent miscarriages, neurologic disorders, osteoporosis, and hyposplenism.^3,4,8

Since CD can manifest with widely varying symptoms, some researchers believe the disease should be classified into 3 categories based on presentation: classic CD, which presents with diarrhea, weight loss, malabsorption, and vitamin deficiency; atypical CD, which presents with minimal GI symptoms but can include anemia, neurologic symptoms, arthritis, or infertility; and asymptomatic CD, which typically displays no symptoms but usually is identified on incidental screening.^3,8,16 Non-celiac gluten sensitivity is a distinct condition in which the body reacts adversely to gluten; it is not an autoimmune disease with an inflammatory response.

Order serologic testing 
for at-risk patients


Because CD remains underdiagnosed,¹⁶ taking a thorough family history and dietary history and making sure to at least consider CD as a part of a differential diagnosis is important. Although population-based screening has been proposed, its benefits and cost-effectiveness remain unproven. As a result, serologic testing of at-risk groups—individuals with conditions known to be associated with CD—remains the current standard.³ The TABLE lists groups for whom serologic testing for CD is indicated.^16,17

In addition, the American College of Gastroenterology (ACG) and the North American Society for Pediatric Gastroenterology, Hepatology, and Nutrition (NASPGHAN) provide guidance on the diagnosis and treatment of adults and children with CD. (An ACG diagnostic algorithm is available at http://www.nature.com/ajg/journal/v108/n5/pdf/ajg201379a.pdf.)

Adults. For patients who are consuming a diet that includes gluten and have symptoms that suggest CD, the ACG guidelines recommend initial testing for IgA tissue transglutaminase (tTG) antibodies.¹⁶ The IgA tTG has a sensitivity and specificity >95%.¹⁶ An alternative test, the IgA endomysial (IgA EMA) test, has similar sensitivity but is time-consuming and its accuracy depends on the experience and skill of the laboratory technician. A negative result for either test has a high negative predictive value for CD.^3,16

IgA deficiency is much more common in patients with CD than in the general population and can result in a false negative test for tTG and EMA. Therefore, consider taking a baseline IgA measurement first. If the patient has an IgA deficiency, the test you’ll use next will change: The preferred test for CD is either immunoglobulin G (IgG) tTG or IgG deamidated gliadin peptides (DGP).^3,16

If a patient is already gluten-free... To rule out CD in patients who are already consuming a gluten-free diet, order HLA-DQ2 and HLA-DQ8 testing because these markers have a specificity >99%; if the HLA test is negative, the disease is excluded.^8,16

Children. NASPGHAN recommends taking a baseline IgA measurement in children at risk for CD and then testing for IgA tTG antibodies, but not until patients are 3 years old and have been on a diet that includes gluten for at least 1 year.¹⁷ Repeat testing at a later date it is recommended for those with negative results because some evidence suggests that in certain patients, later serologic testing will be positive. Alternatively, you may offer HLA testing. If the HLA test is negative, CD can be excluded >99% of the time.

Diagnosis usually is confirmed 
by intestinal biopsy


Positive results on serologic testing should be confirmed with a biopsy of the small bowel; findings characteristic of CD include an increased number of intraepithelial lymphocytes (>25 per 100 enterocytes), elongation of the crypts, and partial to total villous atrophy.⁴ Final confirmation of CD is resolution of symptoms by consuming a gluten-free diet.^3,8

Alternate approaches to confirming the diagnosis. Although intestinal biopsy has long been considered the gold standard for diagnosis of CD, this may change. In 2012, the European Society for Paediatric Gastroenterology, Hepatology, and Nutrition proposed that the biopsy may not be necessary in children with the following 3 characteristics: classic intestinal symptoms of CD, IgA tTG levels >10 times higher than normal, and a positive HLA-DQ2.¹⁸

Catassi and Fasano¹⁹ have proposed shifting from relying on algorithms and intestinal biopsy to a quantitative approach. They suggest using the “4 out of 5” rule, meaning the diagnosis of CD can be confirmed if at least 4 of the following 5 criteria are satisfied: typical CD symptoms, a positive IgA tTG, a positive HLA-DQ2 or -DQ8, celiac enteropathy on small bowel biopsy, and response to a gluten-free diet.¹⁹

The only proven treatment:
 A gluten-free diet

Non-celiac gluten sensitivity is a distinct condition in which the body reacts adversely to gluten; it is not an autoimmune disease with an inflammatory response.Lifelong adherence to a gluten-free diet is the only effective treatment for CD.^14,16 Previously, patients with CD were advised to also avoid oats, but most evidence supports the safety of oats (<2 oz/d), provided there is no cross-contamination with gluten.¹⁴ Adhering to a strict gluten-free diet can be challenging because cereal flours are ubiquitous in western foods, and some foods may be cross-contaminated. The Celiac Disease Foundation (http://www.celiac.org) offers guidance on maintaining a gluten-free diet.

Because avoiding gluten has become popular even among people who don’t have CD, product labeling that includes information on gluten content has become pervasive. However, determining which items contain gluten depends on accurate labeling, a standard that often is not met in many countries; in the United States, such labeling began to be phased in starting in July 2014.²⁰ As a result, CD patients may unwittingly be exposed to gluten over the long term, which can result in greater morbidity and mortality. Unless a food is labeled “gluten-free," it is best to check with the manufacturer.

Compliance with a gluten-free diet can be monitored by following IgA tTG titers every 1 to 2 years, as these values normalize after a patient has been adhering to the diet for 6 to 24 months.^3,16

In addition to lifelong adherence to a gluten-free diet, a National Institutes of Health Consensus Development Conference recommended that management of patients with CD should include²¹:

• consultation with a skilled dietitian

• education about the disease
• 
continuous long-term follow-up by a multidisciplinary team
• 
identification and follow-up of abnormalities found at baseline, such as abnormal liver function test results

• treatment of nutritional deficiencies.

The ACG also recommends that CD patients receive a dual energy x-ray scan for follow-up of osteopenia and a pneumococcal vaccine because functional hyposplenism is associated with CD, and pneumonia is a common complication of hyposplenism.^16,18

Should you recommend a gluten-free diet for other patients? Because avoiding gluten is now popular and many gluten-free products are marketed as “health food,” physicians may be reluctant to recommend a gluten-free diet for patients who have vague abdominal symptoms but negative CD test results. Despite the current popularity of “going gluten-free,” the reality is that in addition to CD, many other diseases may be helped by a gluten-free diet, such as dermatitis herpetiformis, irritable bowel syndrome, and neurologic diseases such as gluten-sensitive ataxia.¹⁹ In the end, whether to adopt a gluten-free diet is a decision that you and your patient will need to make together.

Researchers are searching 
for additional treatments


Because many patients find it difficult to adhere to a gluten-free diet, researchers are investigating several alternative treatments, including a derivative from cholera toxin that inhibits the opening of intestinal epithelial junctions, thereby reducing the resultant inflammatory response, and a desensitizing vaccine.^19,22,23 Another intriguing approach involves using the parasite Necator americanus to modulate the immune response to gluten.18 Finally, certain infant feeding practices, including breastfeeding and delaying introduction of gluten to the diet, may minimize the risk of developing CD. (See "A link between infant feeding practices and the risk of CD?" above.^22-25)

CORRESPONDENCE
Patrick T. Dowling, MD, MPH, Department of Family Medicine, 50-078 Center for Health Sciences, David Geffen School of Medicine at UCLA, Los Angeles, CA 90095-1683; pdowling@mednet.ucla.edu

CASE 1 › A 28-year-old woman (G0P0) came to our office with recurrent episodes of postprandial epigastric and right upper quadrant pain. Upper and lower endoscopy, sonography, body imaging, and laboratory tests were normal. A biliary nuclear scan showed an ejection fraction (EF) of 95%; normal is >35%. We made a diagnosis of biliary dyskinesia (BD) and recommended a laparoscopic cholecystectomy. The patient underwent this procedure and her pain was relieved. She has been much improved for 2 years, although she has since been diagnosed with an autoimmune disorder.

CASE 2 › A 21-year-old woman with right upper quadrant, postprandial, colicky pain presented to the emergency department. The episode lasted approximately 30 minutes and was followed by residual soreness. This episode was one of several that had been increasing in frequency and intensity. A sonogram showed a normal gallbladder and common duct. All laboratory tests were normal. She improved and was discharged. Outpatient evaluation included body imaging and endoscopy, which were negative. A hepatobiliary (HIDA) scan revealed an EF of 90%, and the scan reproduced her symptoms.

We diagnosed BD in this patient. After reviewing the risks and benefits of cholecystectomy, the patient consented to the procedure. She has been asymptomatic for 2 years.

Family physicians often are the first to evaluate patients with recurrent biliary colic. Biliary colic without gallstones—also known as BD or acalculous cholecystitis—is a functional disorder of the gallbladder or bile duct. Approximately 8% of men and 21% of women with biliary pain do not have gallstones.^1-5

BD has been successfully treated with cholecystectomy. Physicians typically have viewed cholecystectomy as being effective primarily for patients with biliary pain who have a low EF (<35%).^2-4 However, recent studies and our experience with cholecystectomy in these 2 patients with high EFs suggest that EF is only one of several factors to consider when deciding whether cholecystectomy might be appropriate for a given patient.

Which patients are most likely
 to benefit from cholecystectomy?

BD is a diagnosis of exclusion, considered when other upper abdominal disorders are eliminated. To receive a diagnosis of BD, patients must meet the Rome III criteria (TABLE).²

Before the advent of oral cholecystography in the 1920s, biliary disease was a clinical diagnosis confirmed by examination of the excised gallbladder.⁶ In 2 large studies conducted before cholecystography was in common use, researchers noted improvement in 75% to 85% of BD patients after cholecystectomy.^7,8 Several years later, with the benefit, of cholecystography, Mackey⁹ reported similar improvement rates among patients with BD who underwent cholecystectomy.

Because cholecystectomy has relieved biliary dyskinesia pain in patients with a wide range
 of ejection fractions, the utility of nuclear scans for such patients
 has been questioned.Cholecystography has largely been replaced with HIDA scanning, which provides an objective measure of EF. Although some studies have suggested low EFs may predict which patients will benefit from cholecystectomy, others have suggested this value doesn’t tell the whole story.^2,4,10,11 In some studies, patients who had biliary symptoms and a low EF (<35%) were found to be most likely to experience relief after cholecystectomy.^2,4 More recently, in a chart review, DuCoin et al¹⁰ found that of 19 BD patients with an EF >35% who underwent cholecystectomy, 17 had complete symptom resolution, one had partial resolution, and one was unchanged. Only one abstract of a study of cholecystectomy for BD patients with a high EF (>80%) has been published.¹¹ Of 28 patients who received cholecystectomy, 22 were asymptomatic after cholecystectomy and 5 others improved.¹¹

Other tests to consider. A cholecystokinin infusion without a scan has been used to reproduce biliary colic; some physicians consider this to be diagnostic of BD and sufficient for cholecystectomy.¹² Others have advocated endoscopic injection of botulinum into the sphincter of Oddi to differentiate pain arising from the sphincter of Oddi from pain in the gallbladder.^5,13 If symptoms are relieved by this injection, an endoscopic biliary sphincterotomy—cutting of the biliary sphincter—is done. Cholecystectomy is reserved for patients whose pain is not relieved by botulinum. In an initial report, 25 BD patients received botulinum injections into the sphincter of Oddi; of the 11 whose pain was relieved by this injection, 10 underwent endoscopic biliary sphincterotomy, and pain resolved for all of these patients.¹³

Why we chose cholecystectomy
 for our patients


Despite a plethora of tests available to visualize and assess gallbladder and bile duct function, clinical assessment of BD by experienced physicians may be sufficient to determine which BD patients will benefit from cholecystectomy. In the cases we report on here, each patient had a high EF, but both met Rome III criteria and were experiencing clinically significant pain. Also, for both patients, a cholecystokinin infusion administered to calculate EF reproduced their pain. This clinical picture led us to recommend laparoscopic cholecystectomy, which ultimately relieved their symptoms.

CORRESPONDENCE
Mazen Iskandar, MD, The Pancreas and Biliary Center of New York, Beth Israel Medical Center, 350 East 17th Street, 16 Baird Hall, New York, NY 10010; miskandar@chpnet.org

CASE 1 › A 28-year-old woman (G0P0) came to our office with recurrent episodes of postprandial epigastric and right upper quadrant pain. Upper and lower endoscopy, sonography, body imaging, and laboratory tests were normal. A biliary nuclear scan showed an ejection fraction (EF) of 95%; normal is >35%. We made a diagnosis of biliary dyskinesia (BD) and recommended a laparoscopic cholecystectomy. The patient underwent this procedure and her pain was relieved. She has been much improved for 2 years, although she has since been diagnosed with an autoimmune disorder.

CASE 2 › A 21-year-old woman with right upper quadrant, postprandial, colicky pain presented to the emergency department. The episode lasted approximately 30 minutes and was followed by residual soreness. This episode was one of several that had been increasing in frequency and intensity. A sonogram showed a normal gallbladder and common duct. All laboratory tests were normal. She improved and was discharged. Outpatient evaluation included body imaging and endoscopy, which were negative. A hepatobiliary (HIDA) scan revealed an EF of 90%, and the scan reproduced her symptoms.

We diagnosed BD in this patient. After reviewing the risks and benefits of cholecystectomy, the patient consented to the procedure. She has been asymptomatic for 2 years.

Family physicians often are the first to evaluate patients with recurrent biliary colic. Biliary colic without gallstones—also known as BD or acalculous cholecystitis—is a functional disorder of the gallbladder or bile duct. Approximately 8% of men and 21% of women with biliary pain do not have gallstones.^1-5

BD has been successfully treated with cholecystectomy. Physicians typically have viewed cholecystectomy as being effective primarily for patients with biliary pain who have a low EF (<35%).^2-4 However, recent studies and our experience with cholecystectomy in these 2 patients with high EFs suggest that EF is only one of several factors to consider when deciding whether cholecystectomy might be appropriate for a given patient.

Which patients are most likely
 to benefit from cholecystectomy?

BD is a diagnosis of exclusion, considered when other upper abdominal disorders are eliminated. To receive a diagnosis of BD, patients must meet the Rome III criteria (TABLE).²

Before the advent of oral cholecystography in the 1920s, biliary disease was a clinical diagnosis confirmed by examination of the excised gallbladder.⁶ In 2 large studies conducted before cholecystography was in common use, researchers noted improvement in 75% to 85% of BD patients after cholecystectomy.^7,8 Several years later, with the benefit, of cholecystography, Mackey⁹ reported similar improvement rates among patients with BD who underwent cholecystectomy.

Because cholecystectomy has relieved biliary dyskinesia pain in patients with a wide range
 of ejection fractions, the utility of nuclear scans for such patients
 has been questioned.Cholecystography has largely been replaced with HIDA scanning, which provides an objective measure of EF. Although some studies have suggested low EFs may predict which patients will benefit from cholecystectomy, others have suggested this value doesn’t tell the whole story.^2,4,10,11 In some studies, patients who had biliary symptoms and a low EF (<35%) were found to be most likely to experience relief after cholecystectomy.^2,4 More recently, in a chart review, DuCoin et al¹⁰ found that of 19 BD patients with an EF >35% who underwent cholecystectomy, 17 had complete symptom resolution, one had partial resolution, and one was unchanged. Only one abstract of a study of cholecystectomy for BD patients with a high EF (>80%) has been published.¹¹ Of 28 patients who received cholecystectomy, 22 were asymptomatic after cholecystectomy and 5 others improved.¹¹

Other tests to consider. A cholecystokinin infusion without a scan has been used to reproduce biliary colic; some physicians consider this to be diagnostic of BD and sufficient for cholecystectomy.¹² Others have advocated endoscopic injection of botulinum into the sphincter of Oddi to differentiate pain arising from the sphincter of Oddi from pain in the gallbladder.^5,13 If symptoms are relieved by this injection, an endoscopic biliary sphincterotomy—cutting of the biliary sphincter—is done. Cholecystectomy is reserved for patients whose pain is not relieved by botulinum. In an initial report, 25 BD patients received botulinum injections into the sphincter of Oddi; of the 11 whose pain was relieved by this injection, 10 underwent endoscopic biliary sphincterotomy, and pain resolved for all of these patients.¹³

Why we chose cholecystectomy
 for our patients


Despite a plethora of tests available to visualize and assess gallbladder and bile duct function, clinical assessment of BD by experienced physicians may be sufficient to determine which BD patients will benefit from cholecystectomy. In the cases we report on here, each patient had a high EF, but both met Rome III criteria and were experiencing clinically significant pain. Also, for both patients, a cholecystokinin infusion administered to calculate EF reproduced their pain. This clinical picture led us to recommend laparoscopic cholecystectomy, which ultimately relieved their symptoms.

CORRESPONDENCE
Mazen Iskandar, MD, The Pancreas and Biliary Center of New York, Beth Israel Medical Center, 350 East 17th Street, 16 Baird Hall, New York, NY 10010; miskandar@chpnet.org

CASE 1 › A 28-year-old woman (G0P0) came to our office with recurrent episodes of postprandial epigastric and right upper quadrant pain. Upper and lower endoscopy, sonography, body imaging, and laboratory tests were normal. A biliary nuclear scan showed an ejection fraction (EF) of 95%; normal is >35%. We made a diagnosis of biliary dyskinesia (BD) and recommended a laparoscopic cholecystectomy. The patient underwent this procedure and her pain was relieved. She has been much improved for 2 years, although she has since been diagnosed with an autoimmune disorder.

CASE 2 › A 21-year-old woman with right upper quadrant, postprandial, colicky pain presented to the emergency department. The episode lasted approximately 30 minutes and was followed by residual soreness. This episode was one of several that had been increasing in frequency and intensity. A sonogram showed a normal gallbladder and common duct. All laboratory tests were normal. She improved and was discharged. Outpatient evaluation included body imaging and endoscopy, which were negative. A hepatobiliary (HIDA) scan revealed an EF of 90%, and the scan reproduced her symptoms.

We diagnosed BD in this patient. After reviewing the risks and benefits of cholecystectomy, the patient consented to the procedure. She has been asymptomatic for 2 years.

Family physicians often are the first to evaluate patients with recurrent biliary colic. Biliary colic without gallstones—also known as BD or acalculous cholecystitis—is a functional disorder of the gallbladder or bile duct. Approximately 8% of men and 21% of women with biliary pain do not have gallstones.^1-5

BD has been successfully treated with cholecystectomy. Physicians typically have viewed cholecystectomy as being effective primarily for patients with biliary pain who have a low EF (<35%).^2-4 However, recent studies and our experience with cholecystectomy in these 2 patients with high EFs suggest that EF is only one of several factors to consider when deciding whether cholecystectomy might be appropriate for a given patient.

Which patients are most likely
 to benefit from cholecystectomy?

BD is a diagnosis of exclusion, considered when other upper abdominal disorders are eliminated. To receive a diagnosis of BD, patients must meet the Rome III criteria (TABLE).²

Before the advent of oral cholecystography in the 1920s, biliary disease was a clinical diagnosis confirmed by examination of the excised gallbladder.⁶ In 2 large studies conducted before cholecystography was in common use, researchers noted improvement in 75% to 85% of BD patients after cholecystectomy.^7,8 Several years later, with the benefit, of cholecystography, Mackey⁹ reported similar improvement rates among patients with BD who underwent cholecystectomy.

Because cholecystectomy has relieved biliary dyskinesia pain in patients with a wide range
 of ejection fractions, the utility of nuclear scans for such patients
 has been questioned.Cholecystography has largely been replaced with HIDA scanning, which provides an objective measure of EF. Although some studies have suggested low EFs may predict which patients will benefit from cholecystectomy, others have suggested this value doesn’t tell the whole story.^2,4,10,11 In some studies, patients who had biliary symptoms and a low EF (<35%) were found to be most likely to experience relief after cholecystectomy.^2,4 More recently, in a chart review, DuCoin et al¹⁰ found that of 19 BD patients with an EF >35% who underwent cholecystectomy, 17 had complete symptom resolution, one had partial resolution, and one was unchanged. Only one abstract of a study of cholecystectomy for BD patients with a high EF (>80%) has been published.¹¹ Of 28 patients who received cholecystectomy, 22 were asymptomatic after cholecystectomy and 5 others improved.¹¹

Other tests to consider. A cholecystokinin infusion without a scan has been used to reproduce biliary colic; some physicians consider this to be diagnostic of BD and sufficient for cholecystectomy.¹² Others have advocated endoscopic injection of botulinum into the sphincter of Oddi to differentiate pain arising from the sphincter of Oddi from pain in the gallbladder.^5,13 If symptoms are relieved by this injection, an endoscopic biliary sphincterotomy—cutting of the biliary sphincter—is done. Cholecystectomy is reserved for patients whose pain is not relieved by botulinum. In an initial report, 25 BD patients received botulinum injections into the sphincter of Oddi; of the 11 whose pain was relieved by this injection, 10 underwent endoscopic biliary sphincterotomy, and pain resolved for all of these patients.¹³

Why we chose cholecystectomy
 for our patients


Despite a plethora of tests available to visualize and assess gallbladder and bile duct function, clinical assessment of BD by experienced physicians may be sufficient to determine which BD patients will benefit from cholecystectomy. In the cases we report on here, each patient had a high EF, but both met Rome III criteria and were experiencing clinically significant pain. Also, for both patients, a cholecystokinin infusion administered to calculate EF reproduced their pain. This clinical picture led us to recommend laparoscopic cholecystectomy, which ultimately relieved their symptoms.

CORRESPONDENCE
Mazen Iskandar, MD, The Pancreas and Biliary Center of New York, Beth Israel Medical Center, 350 East 17th Street, 16 Baird Hall, New York, NY 10010; miskandar@chpnet.org

Failure to spot CHF leads to heart transplant

A 49-YEAR-OLD MAN SOUGHT TREATMENT AT AN URGENT CARE FACILITY after having shortness of breath every morning for 2 weeks. His heart rate was 119 beats/min, his blood pressure was 170/101 mm Hg, and he did not have chest pain. An electrocardiogram (EKG) was abnormal and chest x-ray showed fluid in the lung. The patient was diagnosed with pneumonia, prescribed antibiotics, and told to follow up with his physician. A follow-up chest x-ray 2 weeks later showed an enlarged heart and more fluid in the lung. A computed tomography scan indicated congestive heart failure and an EKG showed signs of a heart attack. The patient underwent a heart transplant and requires immunosuppressants.

In newborns, the differential diagnosis for shortness of breath widens to include infection. PLAINTIFF'S CLAIM If the physician at the urgent care facility had noticed the patient’s enlarged heart, there would have been less heart damage, and the patient might have required a bypass, rather than a transplant.


THE DEFENSE No information about the defense is available.


VERDICT $1 million New Jersey verdict.

COMMENT When evaluating shortness of breath, always think lungs and heart until you have a definite diagnosis. Remember that neurological disease can present with shortness of breath, too. Consider amyotrophic lateral sclerosis, Guillain-Barré syndrome, and myasthenia gravis.

Infant suffers brain injury after delayed lab results

PARENTS BROUGHT THEIR 2-WEEK-OLD DAUGHTER TO THE EMERGENCY DEPARTMENT (ED) after she had missed several feedings and was short of breath. The ED physician ordered blood tests, but discharged the patient before receiving the results and told the parents to follow up with the infant’s pediatrician. Blood work subsequently revealed that the child had a Group B streptococcus infection, but by the time these results were communicated to the parents and treatment had begun, the infant had developed meningitis. She suffered brain injury, and was diagnosed with cerebral palsy.


PLAINTIFF'S CLAIM There was a delay in the diagnosis and treatment of the infant. Blood test results showing a bacterial infection were available the morning after discharge, but instead of notifying the parents, an additional blood culture was ordered to determine the type of bacteria present. The parents were then contacted 6 hours after the bacteria was identified as Group B streptococcus.

THE DEFENSE The defendants denied any negligence, although a nurse who cared for the infant claimed she had expressed concerns about the decision to discharge the patient.

VERDICT $7.15 million Maryland verdict.

COMMENT In newborns, the differential diagnosis for shortness of breath widens to include infection. In this case, I suspect the problem was a lack of tight follow-up, which can lead to bad outcomes—especially in newborns.

Failure to spot CHF leads to heart transplant

A 49-YEAR-OLD MAN SOUGHT TREATMENT AT AN URGENT CARE FACILITY after having shortness of breath every morning for 2 weeks. His heart rate was 119 beats/min, his blood pressure was 170/101 mm Hg, and he did not have chest pain. An electrocardiogram (EKG) was abnormal and chest x-ray showed fluid in the lung. The patient was diagnosed with pneumonia, prescribed antibiotics, and told to follow up with his physician. A follow-up chest x-ray 2 weeks later showed an enlarged heart and more fluid in the lung. A computed tomography scan indicated congestive heart failure and an EKG showed signs of a heart attack. The patient underwent a heart transplant and requires immunosuppressants.

In newborns, the differential diagnosis for shortness of breath widens to include infection. PLAINTIFF'S CLAIM If the physician at the urgent care facility had noticed the patient’s enlarged heart, there would have been less heart damage, and the patient might have required a bypass, rather than a transplant.


THE DEFENSE No information about the defense is available.


VERDICT $1 million New Jersey verdict.

COMMENT When evaluating shortness of breath, always think lungs and heart until you have a definite diagnosis. Remember that neurological disease can present with shortness of breath, too. Consider amyotrophic lateral sclerosis, Guillain-Barré syndrome, and myasthenia gravis.

Infant suffers brain injury after delayed lab results

PARENTS BROUGHT THEIR 2-WEEK-OLD DAUGHTER TO THE EMERGENCY DEPARTMENT (ED) after she had missed several feedings and was short of breath. The ED physician ordered blood tests, but discharged the patient before receiving the results and told the parents to follow up with the infant’s pediatrician. Blood work subsequently revealed that the child had a Group B streptococcus infection, but by the time these results were communicated to the parents and treatment had begun, the infant had developed meningitis. She suffered brain injury, and was diagnosed with cerebral palsy.


PLAINTIFF'S CLAIM There was a delay in the diagnosis and treatment of the infant. Blood test results showing a bacterial infection were available the morning after discharge, but instead of notifying the parents, an additional blood culture was ordered to determine the type of bacteria present. The parents were then contacted 6 hours after the bacteria was identified as Group B streptococcus.

THE DEFENSE The defendants denied any negligence, although a nurse who cared for the infant claimed she had expressed concerns about the decision to discharge the patient.

VERDICT $7.15 million Maryland verdict.

COMMENT In newborns, the differential diagnosis for shortness of breath widens to include infection. In this case, I suspect the problem was a lack of tight follow-up, which can lead to bad outcomes—especially in newborns.

Failure to spot CHF leads to heart transplant

A 49-YEAR-OLD MAN SOUGHT TREATMENT AT AN URGENT CARE FACILITY after having shortness of breath every morning for 2 weeks. His heart rate was 119 beats/min, his blood pressure was 170/101 mm Hg, and he did not have chest pain. An electrocardiogram (EKG) was abnormal and chest x-ray showed fluid in the lung. The patient was diagnosed with pneumonia, prescribed antibiotics, and told to follow up with his physician. A follow-up chest x-ray 2 weeks later showed an enlarged heart and more fluid in the lung. A computed tomography scan indicated congestive heart failure and an EKG showed signs of a heart attack. The patient underwent a heart transplant and requires immunosuppressants.

In newborns, the differential diagnosis for shortness of breath widens to include infection. PLAINTIFF'S CLAIM If the physician at the urgent care facility had noticed the patient’s enlarged heart, there would have been less heart damage, and the patient might have required a bypass, rather than a transplant.


THE DEFENSE No information about the defense is available.


VERDICT $1 million New Jersey verdict.

COMMENT When evaluating shortness of breath, always think lungs and heart until you have a definite diagnosis. Remember that neurological disease can present with shortness of breath, too. Consider amyotrophic lateral sclerosis, Guillain-Barré syndrome, and myasthenia gravis.

Infant suffers brain injury after delayed lab results

PARENTS BROUGHT THEIR 2-WEEK-OLD DAUGHTER TO THE EMERGENCY DEPARTMENT (ED) after she had missed several feedings and was short of breath. The ED physician ordered blood tests, but discharged the patient before receiving the results and told the parents to follow up with the infant’s pediatrician. Blood work subsequently revealed that the child had a Group B streptococcus infection, but by the time these results were communicated to the parents and treatment had begun, the infant had developed meningitis. She suffered brain injury, and was diagnosed with cerebral palsy.


PLAINTIFF'S CLAIM There was a delay in the diagnosis and treatment of the infant. Blood test results showing a bacterial infection were available the morning after discharge, but instead of notifying the parents, an additional blood culture was ordered to determine the type of bacteria present. The parents were then contacted 6 hours after the bacteria was identified as Group B streptococcus.

THE DEFENSE The defendants denied any negligence, although a nurse who cared for the infant claimed she had expressed concerns about the decision to discharge the patient.

VERDICT $7.15 million Maryland verdict.

COMMENT In newborns, the differential diagnosis for shortness of breath widens to include infection. In this case, I suspect the problem was a lack of tight follow-up, which can lead to bad outcomes—especially in newborns.

ABSTRACT

Purpose We undertook this study to explore the factors associated with differences between patients’ stated main reasons for outpatient visits and physicians’ main concerns at those same visits.

Methods This cross-sectional, mixed-methods study examined 192 outpatient visits with 4 physicians at 4 diverse primary care practices. During each visit, participating physicians elicited the patient’s main reason for the visit. Immediately after each visit, physicians documented 1) their understanding of the patient’s stated reason and 2) their main concern for the patient during that visit, and 3) assessed the extent of their alignment with the patient’s reason for visit. We assessed bivariate and multivariable associations of patient and visit characteristics with alignment, and further examined cases with unaligned physician-patient priorities to identify patterns.

Results In 69% of visits, the patient’s stated reason for the visit was completely aligned with the physician’s main concern. In 12% of visits, we observed totally unaligned priorities; 19% were only partially aligned. Uninsured or publicly-insured patients and visits with more problems addressed were less likely to be fully aligned. In many visits with unaligned priorities, patients’ stated reason for the visit was a self-limiting, symptomatic concern while physicians prioritized potentially dangerous asymptomatic conditions or ill-managed chronic conditions.

Conclusions In diverse family medicine practices, lack of alignment between physician and patient visit priorities reflects differing prioritization processes. Patients presenting with concerns unaligned with their physician’s priorities may require more time or different approaches to ensure the relevance and patient-centeredness of their care. These findings may inform the design of systems of care that promote mindful attention to patients’ priorities while addressing medically urgent or preventive services delivery.

T
oday’s family physicians must balance patient’s acute concerns with chronic disease management, health promotion, and disease prevention. It’s not easy. As the content of outpatient visits expands and available time contracts,¹ patients’, clinicians’, and payers’ agendas compete for attention. From a patient experience perspective, the health care encounter may seem diminished when guideline-driven agendas championing chronic disease management and preventive service delivery appear to take precedence over their personal concerns.^2-4

In the matter of physician-patient alignment of visit priorities, prior research^5,6 inadequately reflects current practice realities such as increased time pressure,⁷ greater chronic disease prevalence,⁸ growing expectations for preventive care,¹ and increasing physician proactivity in longitudinal care.^9-12 With so much to do and so little time and mounting pressure to deliver patient-centered care and patient satisfaction, it would be helpful to have a better understanding of how often and with whom physicians choose to depart from a patient’s explicitly stated reasons for a visit and instead prioritize other concerns.

Patient 
factors that were independently associated with less than fully aligned visit priorities included having more problems to address and being publicly insured or uninsured. We sought to examine alignment between patients’ stated main reasons for a visit as understood by the physician and the physician’s main concern during that same visit. Using a diverse sample of family physician-researchers to serve as data collectors and analysts, this study aimed to identify patient and visit characteristics associated with differing physician-patient visit priorities.

METHODS

Study design and sample

Four family physicians participating in a research fellowship undertook this cross- sectional descriptive study of a sample of their outpatient encounters. Each physician’s practice was unique: a free clinic, an inner city family practice within a teaching hospital, a geriatric home visit practice, and a suburban pediatric practice. Using the card study method pioneered by the Ambulatory Sentinel Practice Network,^13,14 physicians collected observational and reflective data on a sample of 50 consecutive patients seen at his or her primary care practice. The University Hospitals Case Medical Center Institutional Review Board approved the study protocol.

Measures

Immediately following each patient visit, physicians recorded on a standardized data card the patient’s characteristics, visit characteristics, the reason for the visit provided to the office staff, the reason for visit reported by the patient at the beginning of the visit, and the physician’s own main concern for the patient during the visit. Patient characteristics included gender, age, race, type of insurance, and number of chronic conditions on the problem list. Visit characteristics included the total number of problems addressed, whether a second person (eg, family member, caregiver) was present in the exam room during the visit, and visit duration. To elicit the patient’s reason for the visit, physicians systematically asked, “How can I help you today?” If more than one problem was elicited from this prompt, the concern expressed as most important by the patient was noted.

In completing the data card, physicians also reflected on the extent to which the patient’s stated reason for the visit aligned with their own main concern for the patient during the visit, rating the relationship between their respective priorities as fully aligned, partially aligned, or totally unaligned. Visits were considered fully aligned if the patient’s reason for the visit and the physician’s concern were the same, or if the patient’s expressed concern was determined to be a symptom related to the physician’s main concern. Partial alignment occurred when the patient’s concern was shared by the physician but was not the physician’s main concern. Visit priorities were rated as totally unaligned if patient and physician concerns were different and determined not to be medically related.

Data analysis

We computed descriptive statistics and performed bivariate tests of association between physician-patient alignment of priorities and patient/visit characteristics, using chi-square statistics for categorical variables and analysis of variance (ANOVA) for continuous variables. We used logistic regression analysis to identify characteristics independently associated with either partially aligned or totally unaligned visit priorities, vs fully aligned priorities.

To better understand the unaligned cases, the study team examined all of the data cards identified as either only partially aligned or totally unaligned and categorized the nature of the disparate prioritization in each case. Study team members (PT, AW, MR, and PDG) individually reviewed and sorted the cards into similar observed patterns and wrote brief descriptions of those patterns. The group refined the preliminary descriptions of the overarching patterns and selected several cases to represent the patterns.

RESULTS

Quantitative findings


We collected data on 192 patients (TABLE 1). Median patient age was 47. Median number of prescription medications and chronic medical conditions were 3 and 2, respectively. Most patients (80%) primarily spoke English, 58% were female, and 59% were Caucasian. Insurance status was equally distributed among patients: 33% held private insurance, 33% had public insurance, and 34% were uninsured.

Physician-perceived time pressure during a visit and lack of an established patient-physician relationship did not influence priority alignment. The majority of cases showed physician- patient priorities that were fully aligned (69%); only 12% were totally unaligned. TABLE 1 shows the associations between level of alignment and patient and visit characteristics. Patients whose expressed reason for their visit was less than fully aligned with their physicians’ primary concern were more likely to be older, with more chronic conditions, of Hispanic ethnicity, and either publicly insured or uninsured. Patients with some degree of unaligned concerns also tended to have greater numbers of prescribed medications and had a primary language other than English. Unaligned visit priorities were also observed more often when greater numbers of problems were addressed during the visit. Patients familiar to the physician were also more likely to express a reason for their visit that differed in some way from the primary concern of their family physician.

Multivariable logistic regression analyses showed that having more problems addressed during a visit and being publicly insured or uninsured (compared with having private insurance) were each independently associated with less than fully aligned visit priorities (indicated by § in TABLE 1).

Qualitative findings

Cases classified as partially or totally unaligned showed 2 main patterns that reflect differences in the ways that patients and physicians prioritize problems. First, when physician priorities were unaligned with the patient’s stated reason for a visit, physicians typically focused on treatable, asymptomatic conditions that could lead to long-term morbidity and mortality or, absent immediate attention, to a dangerous event. Examples include chronic disease management (eg, hypertension, diabetes, asthma) and disease prevention (eg, smoking cessation, cholesterol level monitoring, cardiac disease risk reduction). Second, in those visits that were unaligned, patients tended to show more initial concern about symptomatic problems (eg, skin rash, upper respiratory infection) that would likely resolve with minimal or no medical intervention. Most of these conditions were determined by the physician to be benign, self-limited, or not worrisome (TABLE 2).

DISCUSSION

This study examined the alignment between a patient’s stated reason for a visit and the physician’s main concern. Physician-patient alignment was common, with 69% of visits showing full alignment between the patient’s reason for the visit and the physician’s main concern. While this rate is higher than that seen in prior studies,⁵ our method for data collection uniquely reflected how information actually is elicited and received during outpatient encounters, which likely yielded more accurate results. Also, in contrast to past approaches, our study equated patients’ symptoms to the underlying diagnosis for purposes of determining alignment.

Similar to previous studies, alignment between patient and physician concerns was less likely when the number of items addressed during the visit increased.^5,6 Interestingly, physician-perceived time pressure during the visit and lack of an established patient-physician relationship were not observed to influence alignment. When time is limited, physicians may choose to prioritize the patient’s verbalized concern to avoid more complex negotiations about the agenda. In addition, during an encounter with an unfamiliar patient, prevention and chronic illness care may be deferred until the relationship becomes established.^15-18

Patient insurance status was also observed to influence alignment. This finding could reflect socioeconomic or access issues among uninsured or publicly-insured patients that contribute to more complex visits.^19-21 To shed further light on this finding, further research is needed that includes indicators of education or income and that separately examines Medicaid vs Medicare enrollees.

Possible study weaknesses. A potential weakness of this study is that alignment was rated by the physicians who elicited and recorded their patients’ reasons for the visit. It is possible that the central role of the physician as observer and analyst may have introduced bias and that an outside observer may have come to different conclusions about the extent to which priorities were aligned. However, by placing the physician-researcher at the center of both data collection and analysis, our study method reflects the realistic constraints of limited information on the processes of clinical discovery and sense-making that physicians regularly undertake with their patients, which we believe to be a strength of this study. While the inclusion of pediatric visits in the sample is another possible weakness, we believe that the diverse clinical settings and diverse patient populations potentially strengthen the findings. Replication with different samples of clinicians and patients is needed to assess the robustness of the findings.

Physicians tend to evaluate risk factors for future disease, while patients focus on symptomatically troublesome—though self-limiting— conditions.Root differences in physician and patient perspectives. Collectively, these findings suggest that patients’ and physicians’ differing approaches to prioritization may limit alignment. In general, physicians tend to evaluate the full scope of the patient’s health and risk factors for future disease, while patients more often focus on symptomatically troublesome—though often self-limiting— conditions. Physicians have the knowledge and clinical experience to prioritize or deprioritize patient concerns based on an assessment of long-term risks of morbidity and mortality, yet the future-orientation of treatments and surveillance for insensible conditions (like hypertension) is less likely to align with the immediately painful or worrisome symptoms of patients. This highlights the importance of patient education on chronic disease management and disease prevention, and cooperative agenda-setting. Further work needs to be done to examine the differences in patients’ and physicians’ cognitive processes of prioritization, with the ultimate goal of providing patient-centered care through shared decision-making.

Patients who are less likely to share the physician’s prioritization of their concerns may require more time and effort on the part of the physician to create a mutually acceptable agenda for the visit.Take-home messages for all stakeholders. Amidst growing time pressures and guideline-driven protocols for care, it is important to attend to the intersecting and diverging patient, physician, and payer agendas that drive the content of the visit. Patients who are less likely to share the physician’s prioritization of their concerns—including individuals with no insurance or public insurance and those with multiple medical problems—may require more time and additional effort on the part of the physician to create a mutually acceptable agenda for the visit. Attempts at pay-for-performance should consider patients’ priorities and preferences for care, particularly when those preferences differ from the priorities of physicians or health insurance plans. A more thorough understanding of patient and physician prioritization during primary care visits could potentially guide the organization of outpatient care and inform the mindful physician’s patient-centered practice to maximize patient benefit.

CORRESPONDENCE
Susan A. Flocke, PhD, Family Medicine Research Division, Case Western Reserve University, 11000 Cedar Avenue, Suite 402, Cleveland, OH 44106; susan.flocke@case.edu

ACKNOWLEDGEMENTS
This study was completed as part of the culture of inquiry fellowship, supported by Academic Administrative Units in primary care Grant #D54HP05444 from the Health Resources and Services Administration, US Department of Health and Human Services. Dr. Stange’s time is supported in part by a clinical research professorship from the American Cancer Society.

ABSTRACT

Purpose We undertook this study to explore the factors associated with differences between patients’ stated main reasons for outpatient visits and physicians’ main concerns at those same visits.