Cleveland Clinic Journal of Medicine

The scenario is common: a patient complains of chronic hoarseness, cough, throat-clearing, sore throat, dysphagia, or a lump in the throat and undergoes laryngoscopy. If this test rules out cancer, the patient is given a diagnosis of laryngopharyngeal reflux (LPR), ie, a form of gastroesophageal reflux disease (GERD) in which the stomach contents get all the way up into the pharynx and down into the larynx. A proton pump inhibitor (PPI) is often prescribed, usually twice daily for 2 months.^1–6

In this article, we review the current understanding of the pathophysiology of LPR and evaluate current diagnostic tests and treatment regimens for patients with suspected LPR.

THE PATHOPHYSIOLOGY OF LPR IS POORLY UNDERSTOOD

Transient relaxation of the lower esophageal sphincter

In a study in 10 healthy volunteers, Dent et al⁷ found that the pressure in the lower esophageal sphincter varies considerably over a 12-hour period. Episodes of reflux were not related to low basal (resting) pressure. Rather, 70% to 100% of reflux episodes occurred during random episodes of transient, complete, and inappropriate relaxation of the sphincter that lasted about 5 to 30 seconds. The mechanism of this relaxation is not known but is thought to be related to activation of the vagus nerve, possibly as a consequence of gastric distention.⁸

Gastric, not duodenal products seem to cause the damage

In a study in dogs, Adhami et al⁹ evaluated the possible role of gastric juices (acid and pepsin) vs duodenal juices (bile acids and trypsin) in laryngeal tissue damage. After taking baseline biopsy samples of the larynx, the investigators applied a variety of gastric and duodenal enzymes at varying pH levels (pH 1–7) to the larynxes. After 9 to 12 applications, they took another biopsy and assessed the changes visually and histologically.

At low (ie, acidic) pH levels, pepsin and conjugated bile acids were the most injurious, causing erythema and histologic evidence of inflammation. The authors concluded that gastric and not duodenal substances cause laryngeal injury and that acid-suppressive therapy “should eliminate the injurious potential” of acid reflux.⁹

The larynx is more sensitive than the esophagus

Monitoring of esophageal pH has shown that healthy people can tolerate as many as 50 episodes a day of acid reflux (pH < 4) in the esophagus. However, Koufman¹⁰ found that as few as three episodes of laryngeal reflux per week can cause severe laryngeal inflammation and injury.

Does pepsin deplete buffers, worsening acid damage?

Johnston et al¹¹ took biopsies from a control group of healthy volunteers and from patients diagnosed with LPR. They detected pepsin in the samples from eight of the nine patients with LPR but in none of the controls. Furthermore, the tissue from patients with LPR had low levels of carbonic anhydrase III. The authors hypothesized that pepsin depletes the laryngopharynx of carbonic anhydrase III, and that therefore these tissues cannot produce enough bicarbonate to buffer the gastric acid. Less bicarbonate would mean greater acidity, so the pepsin would remain active and would be more likely to cause cellular damage.¹¹

However, this contention is controversial. What is universally agreed upon is that reflux of gastric or gastroduodenal contents is most likely causing injury, most likely through direct exposure, although indirect effects through vagal mechanisms cannot be ruled out.

CURRENT DIAGNOSTIC TESTS FOR LPR HAVE SHORTCOMINGS

A careful history is important. Many patients report they have sore throat, hoarseness, cough, dysphasia, or chronic throat-clearing.¹³ Factors that may predispose a patient to esophageal reflux should be discussed, eg:

• Tobacco use
• Diet (eg, soda, spicy foods, fatty foods)
• Alcohol use
• Certain drugs (calcium channel blockers, nitrates, steroids).

Up to 50% of patients presenting with extraesophageal symptoms may not have classic reflux symptoms such as heartburn and regur-gitation.¹⁴ However, the existence of “silent reflux” is currently controversial.

Because the key symptoms of LPR are nonspecific, many patients who present to an otolaryngologist undergo laryngoscopy, mainly to rule out malignancy. Once cancer is ruled out, many patients are given a diagnosis of LPR.

Furthermore, Branski et al¹⁷ performed transoral rigid laryngoscopy with videorecording in 100 consecutive patients presenting with a chief complaint of dysphonia. Five board-certified otolaryngologists individually viewed each recording, scored the degree of erythema and edema, and assessed the likelihood that LPR played a role in dysphonia and the severity of the LPR findings. The physicians’ ratings showed considerable interobserver variability. In other words, this study showed that laryngeal findings are often nonspecific and that the laryngoscopic diagnosis of LPR tends to be subjective.¹⁷

The Reflux Finding Score. Concerned by the lack of consistency in the diagnosis of LPR, Belafsky et al¹⁸ created a scoring system for documenting the physical findings and severity of disease on a standardized scale. Their Reflux Finding Score is based on eight laryngoscopic findings: subglottic edema, ventricular edema, erythema, vocal cord edema, diffuse laryngeal edema, hypertrophy of the posterior commissure, granuloma or granulation tissue, and thick endolaryngeal mucus. The total score can range from 0 (best) to 26 (worst).

In 40 patients with LPR confirmed by pH monitoring, the mean score was 11.5, compared with 5.2 in 40 age-matched controls. The authors calculated they could be 95% certain that a person with a score higher than 7 has LPR.¹⁸

However, this diagnostic method has not been validated in a large-scale randomized trial and so has yet to be incorporated into routine otolaryngology practice.

Ambulatory pH monitoring is not so golden for diagnosing LPR

Although pH monitoring was once the gold standard for diagnosing reflux, it has since been shown to be unreliable in patients who have laryngeal symptoms.⁴

How high or low in the esophagus the probe is placed is clearly critical for useful results. ⁴ But the test is subject to variability: different physicians place the probe in different locations, and the probe may shift. Another problem is that reflux may occur during untested periods.¹⁹

A pH of less than 4 in the esophagus had originally been shown to have high sensitivity and specificity,²⁰ but Reichel and Issing²¹ suggested using a pH of less than 5 as the cutoff, which would identify more patients as having LPR. Further trials are needed to more precisely determine the pH threshold for the diagnosis of LPR.

Enthusiasm is waning for pharyngeal pH monitoring

In LPR, it was initially thought that pH monitoring in the pharynx was more accurate than in the distal or proximal esophagus.

Shaker et al²² monitored the pH in the pharynx, proximal esophagus, and distal esophagus in four groups: 14 patients who had both laryngeal signs and symptoms, 12 patients who had laryngeal symptoms only, 16 patients who had GERD but no laryngeal symptoms, and 12 healthy volunteers. They found that pharyngeal reflux was more frequent and in greater quantity in patients with laryngeal signs and symptoms than in the other groups. This study suggested that pharyngeal pH monitoring may be useful in diagnosing LPR in patients who have laryngeal signs and symptoms.

However, hypopharyngeal pH monitoring has several problems. One issue is that, even in this trial, 2 of 12 healthy volunteers had episodes of pharyngeal reflux.²² In other studies, the rate of false-positive results ranged from 7% to 17%.^23,24 Additionally, in 12 previous studies, only 54% of 1,217 patients with suspected LPR had esophageal acid exposure, regardless of where the pH probe was placed.²⁵

More importantly, another study found that patients with pharyngeal reflux documented by pH monitoring were no more likely to respond to acid-suppressive therapy than patients with no documented reflux.²⁶ These findings dampen the enthusiasm for pharyngeal pH monitoring in LPR.

Impedance monitoring on therapy may be useful in refractory cases

Esophageal impedance monitoring, a newer test, uses a catheter that measures electrical resistance (impedance) between different points along the esophagus. Thus, it can detect the reflux of acid and nonacid liquid or gaseous material.

Pritchett et al²⁷ performed esophageal impedance and pH monitoring in 39 patients who were on twice-daily PPI therapy and then evaluated the same patients with wireless pH monitoring while they were off therapy. The most prevalent complaint in the study group was cough (56%), followed by heartburn (18%) and sore throat (10%).

Of the 39 patients, 25 (64%) had normal results on impedance/pH monitoring while on therapy, ruling out reflux. On pH monitoring off therapy, 28 (72%) of the 39 patients had abnormal results; this group included 13 (93%) of the 14 patients who had abnormal results on impedance/pH monitoring while on therapy. The authors recommended on-therapy testing with impedance monitoring in patients with refractory reflux, since it provides more useful clinical information.²⁷ If the results of impedance/pH monitoring are negative in these patients, a diagnosis other than reflux should be considered.

EMPIRIC PPI TREATMENT HAS SHOWN DISAPPOINTING RESULTS

Because laryngoscopy and pH monitoring are not very sensitive or specific for LPR, experts recommend empiric therapy with a PPI twice daily. However, the results have been disappointing when PPIs were compared with placebo in clinical trials.

In a randomized controlled trial,28 we found that patients who had complaints of chronic throat-clearing, cough, globus, sore throat, and hoarseness had a similar response to twice-daily esomeprazole (Nexium) compared with placebo: their primary symptom had resolved by 16 weeks in 14.7% of the esomeprazole group vs 16.0% of the placebo group (P = .799). Similarly, the final findings on laryngoscopy such as edema, erythema, and surface irregularity were not significantly different between groups.

From Qadeer MA, et al. Proton pump inhibitor therapy for suspected GERD-related chronic laryngitis: a meta analysis of randomized controlled trials. Am J Gastroenterol 2006; 101:2646–2654. Used with permission from Nature Publishing Group.

Adding a histamine-2 receptor antagonist is not recommended

Adding a histamine-2 receptor antagonist to PPI therapy has also been considered as a treatment for LPR.

Fackler et al³⁸ studied 16 GERD patients and 18 healthy volunteers to determine if adding ranitidine (Pepcid) to the PPI omeprazole (Prilosec) could improve GERD symptoms. Patients underwent baseline manometry and then gastroesophageal pH monitoring before starting the drugs. They first received omeprazole 20 mg twice daily alone for 2 weeks, and then added ranitidine 300 mg at bedtime. A pH test was done again after the first day of treatment with ranitidine, at the end of 1 week of combination therapy, and after 4 weeks of combination therapy. The combination reduced nocturnal acid breakthrough on day 1; however, due to tolerance to ranitidine, no significant difference in acid suppression was seen after 1 week of therapy. Therefore, this combination is not recommended.

Surgery is not recommended either

Some experts have argued for surgical fundoplication in patients whose symptoms persist despite drug therapy.

Swoger et al³⁹ treated 72 patients who had symptoms consistent with LPR with a PPI for 4 months; 25 patients in this group had less than a 50% improvement despite maximal drug therapy. Ten of these patients underwent surgical fundoplication, and 15 remained on drug therapy alone. At 1 year of follow-up, only one surgical patient (10%) reported improvement in laryngeal symptoms.

In view of this report and prior studies of surgical fundoplication,⁴⁰ surgery is not recommended for patients whose symptoms do not respond to aggressive PPI therapy.

IF A PPI FAILS, LOOK FOR OTHER CAUSES OF SYMPTOMS

Although gastroesophageal reflux and laryngeal signs and symptoms have been associated with one another, this relation may have been overstated, leading to the overdiagnosis of LPR.

The diagnosis of LPR is difficult, as laryngoscopy has high interrater variability and as the results of pH monitoring do not dependably predict who will respond to treatment.

Because PPI therapy is easy and appears to be safe, patients with extraesophageal symptoms thought to be related to reflux should undergo a trial of twice-daily PPI therapy for at least 2 months. If the patient responds to therapy, then tapering to once-daily therapy initially and then to minimal acid suppression to control symptoms would be prudent.

In patients who show no improvement, other causes of symptoms should be explored. Diseases that can mimic LPR include postnasal drip, allergies, sinus inflammation, and various pulmonary diseases. These patients should also be advised to adopt lifestyle modifications—eg, to stop smoking, lose weight, and decrease activities that cause stress on the voice. Surgery is not likely to provide any benefit in this situation. The patient should be tapered off the PPI to make sure no rebound acid reflux occurs.

The scenario is common: a patient complains of chronic hoarseness, cough, throat-clearing, sore throat, dysphagia, or a lump in the throat and undergoes laryngoscopy. If this test rules out cancer, the patient is given a diagnosis of laryngopharyngeal reflux (LPR), ie, a form of gastroesophageal reflux disease (GERD) in which the stomach contents get all the way up into the pharynx and down into the larynx. A proton pump inhibitor (PPI) is often prescribed, usually twice daily for 2 months.^1–6

In this article, we review the current understanding of the pathophysiology of LPR and evaluate current diagnostic tests and treatment regimens for patients with suspected LPR.

THE PATHOPHYSIOLOGY OF LPR IS POORLY UNDERSTOOD

Transient relaxation of the lower esophageal sphincter

In a study in 10 healthy volunteers, Dent et al⁷ found that the pressure in the lower esophageal sphincter varies considerably over a 12-hour period. Episodes of reflux were not related to low basal (resting) pressure. Rather, 70% to 100% of reflux episodes occurred during random episodes of transient, complete, and inappropriate relaxation of the sphincter that lasted about 5 to 30 seconds. The mechanism of this relaxation is not known but is thought to be related to activation of the vagus nerve, possibly as a consequence of gastric distention.⁸

Gastric, not duodenal products seem to cause the damage

In a study in dogs, Adhami et al⁹ evaluated the possible role of gastric juices (acid and pepsin) vs duodenal juices (bile acids and trypsin) in laryngeal tissue damage. After taking baseline biopsy samples of the larynx, the investigators applied a variety of gastric and duodenal enzymes at varying pH levels (pH 1–7) to the larynxes. After 9 to 12 applications, they took another biopsy and assessed the changes visually and histologically.

At low (ie, acidic) pH levels, pepsin and conjugated bile acids were the most injurious, causing erythema and histologic evidence of inflammation. The authors concluded that gastric and not duodenal substances cause laryngeal injury and that acid-suppressive therapy “should eliminate the injurious potential” of acid reflux.⁹

The larynx is more sensitive than the esophagus

Monitoring of esophageal pH has shown that healthy people can tolerate as many as 50 episodes a day of acid reflux (pH < 4) in the esophagus. However, Koufman¹⁰ found that as few as three episodes of laryngeal reflux per week can cause severe laryngeal inflammation and injury.

Does pepsin deplete buffers, worsening acid damage?

Johnston et al¹¹ took biopsies from a control group of healthy volunteers and from patients diagnosed with LPR. They detected pepsin in the samples from eight of the nine patients with LPR but in none of the controls. Furthermore, the tissue from patients with LPR had low levels of carbonic anhydrase III. The authors hypothesized that pepsin depletes the laryngopharynx of carbonic anhydrase III, and that therefore these tissues cannot produce enough bicarbonate to buffer the gastric acid. Less bicarbonate would mean greater acidity, so the pepsin would remain active and would be more likely to cause cellular damage.¹¹

However, this contention is controversial. What is universally agreed upon is that reflux of gastric or gastroduodenal contents is most likely causing injury, most likely through direct exposure, although indirect effects through vagal mechanisms cannot be ruled out.

CURRENT DIAGNOSTIC TESTS FOR LPR HAVE SHORTCOMINGS

A careful history is important. Many patients report they have sore throat, hoarseness, cough, dysphasia, or chronic throat-clearing.¹³ Factors that may predispose a patient to esophageal reflux should be discussed, eg:

• Tobacco use
• Diet (eg, soda, spicy foods, fatty foods)
• Alcohol use
• Certain drugs (calcium channel blockers, nitrates, steroids).

Up to 50% of patients presenting with extraesophageal symptoms may not have classic reflux symptoms such as heartburn and regur-gitation.¹⁴ However, the existence of “silent reflux” is currently controversial.

Because the key symptoms of LPR are nonspecific, many patients who present to an otolaryngologist undergo laryngoscopy, mainly to rule out malignancy. Once cancer is ruled out, many patients are given a diagnosis of LPR.

Furthermore, Branski et al¹⁷ performed transoral rigid laryngoscopy with videorecording in 100 consecutive patients presenting with a chief complaint of dysphonia. Five board-certified otolaryngologists individually viewed each recording, scored the degree of erythema and edema, and assessed the likelihood that LPR played a role in dysphonia and the severity of the LPR findings. The physicians’ ratings showed considerable interobserver variability. In other words, this study showed that laryngeal findings are often nonspecific and that the laryngoscopic diagnosis of LPR tends to be subjective.¹⁷

The Reflux Finding Score. Concerned by the lack of consistency in the diagnosis of LPR, Belafsky et al¹⁸ created a scoring system for documenting the physical findings and severity of disease on a standardized scale. Their Reflux Finding Score is based on eight laryngoscopic findings: subglottic edema, ventricular edema, erythema, vocal cord edema, diffuse laryngeal edema, hypertrophy of the posterior commissure, granuloma or granulation tissue, and thick endolaryngeal mucus. The total score can range from 0 (best) to 26 (worst).

In 40 patients with LPR confirmed by pH monitoring, the mean score was 11.5, compared with 5.2 in 40 age-matched controls. The authors calculated they could be 95% certain that a person with a score higher than 7 has LPR.¹⁸

However, this diagnostic method has not been validated in a large-scale randomized trial and so has yet to be incorporated into routine otolaryngology practice.

Ambulatory pH monitoring is not so golden for diagnosing LPR

Although pH monitoring was once the gold standard for diagnosing reflux, it has since been shown to be unreliable in patients who have laryngeal symptoms.⁴

How high or low in the esophagus the probe is placed is clearly critical for useful results. ⁴ But the test is subject to variability: different physicians place the probe in different locations, and the probe may shift. Another problem is that reflux may occur during untested periods.¹⁹

A pH of less than 4 in the esophagus had originally been shown to have high sensitivity and specificity,²⁰ but Reichel and Issing²¹ suggested using a pH of less than 5 as the cutoff, which would identify more patients as having LPR. Further trials are needed to more precisely determine the pH threshold for the diagnosis of LPR.

Enthusiasm is waning for pharyngeal pH monitoring

In LPR, it was initially thought that pH monitoring in the pharynx was more accurate than in the distal or proximal esophagus.

Shaker et al²² monitored the pH in the pharynx, proximal esophagus, and distal esophagus in four groups: 14 patients who had both laryngeal signs and symptoms, 12 patients who had laryngeal symptoms only, 16 patients who had GERD but no laryngeal symptoms, and 12 healthy volunteers. They found that pharyngeal reflux was more frequent and in greater quantity in patients with laryngeal signs and symptoms than in the other groups. This study suggested that pharyngeal pH monitoring may be useful in diagnosing LPR in patients who have laryngeal signs and symptoms.

However, hypopharyngeal pH monitoring has several problems. One issue is that, even in this trial, 2 of 12 healthy volunteers had episodes of pharyngeal reflux.²² In other studies, the rate of false-positive results ranged from 7% to 17%.^23,24 Additionally, in 12 previous studies, only 54% of 1,217 patients with suspected LPR had esophageal acid exposure, regardless of where the pH probe was placed.²⁵

More importantly, another study found that patients with pharyngeal reflux documented by pH monitoring were no more likely to respond to acid-suppressive therapy than patients with no documented reflux.²⁶ These findings dampen the enthusiasm for pharyngeal pH monitoring in LPR.

Impedance monitoring on therapy may be useful in refractory cases

Esophageal impedance monitoring, a newer test, uses a catheter that measures electrical resistance (impedance) between different points along the esophagus. Thus, it can detect the reflux of acid and nonacid liquid or gaseous material.

Pritchett et al²⁷ performed esophageal impedance and pH monitoring in 39 patients who were on twice-daily PPI therapy and then evaluated the same patients with wireless pH monitoring while they were off therapy. The most prevalent complaint in the study group was cough (56%), followed by heartburn (18%) and sore throat (10%).

Of the 39 patients, 25 (64%) had normal results on impedance/pH monitoring while on therapy, ruling out reflux. On pH monitoring off therapy, 28 (72%) of the 39 patients had abnormal results; this group included 13 (93%) of the 14 patients who had abnormal results on impedance/pH monitoring while on therapy. The authors recommended on-therapy testing with impedance monitoring in patients with refractory reflux, since it provides more useful clinical information.²⁷ If the results of impedance/pH monitoring are negative in these patients, a diagnosis other than reflux should be considered.

EMPIRIC PPI TREATMENT HAS SHOWN DISAPPOINTING RESULTS

Because laryngoscopy and pH monitoring are not very sensitive or specific for LPR, experts recommend empiric therapy with a PPI twice daily. However, the results have been disappointing when PPIs were compared with placebo in clinical trials.

In a randomized controlled trial,28 we found that patients who had complaints of chronic throat-clearing, cough, globus, sore throat, and hoarseness had a similar response to twice-daily esomeprazole (Nexium) compared with placebo: their primary symptom had resolved by 16 weeks in 14.7% of the esomeprazole group vs 16.0% of the placebo group (P = .799). Similarly, the final findings on laryngoscopy such as edema, erythema, and surface irregularity were not significantly different between groups.

From Qadeer MA, et al. Proton pump inhibitor therapy for suspected GERD-related chronic laryngitis: a meta analysis of randomized controlled trials. Am J Gastroenterol 2006; 101:2646–2654. Used with permission from Nature Publishing Group.

Adding a histamine-2 receptor antagonist is not recommended

Adding a histamine-2 receptor antagonist to PPI therapy has also been considered as a treatment for LPR.

Fackler et al³⁸ studied 16 GERD patients and 18 healthy volunteers to determine if adding ranitidine (Pepcid) to the PPI omeprazole (Prilosec) could improve GERD symptoms. Patients underwent baseline manometry and then gastroesophageal pH monitoring before starting the drugs. They first received omeprazole 20 mg twice daily alone for 2 weeks, and then added ranitidine 300 mg at bedtime. A pH test was done again after the first day of treatment with ranitidine, at the end of 1 week of combination therapy, and after 4 weeks of combination therapy. The combination reduced nocturnal acid breakthrough on day 1; however, due to tolerance to ranitidine, no significant difference in acid suppression was seen after 1 week of therapy. Therefore, this combination is not recommended.

Surgery is not recommended either

Some experts have argued for surgical fundoplication in patients whose symptoms persist despite drug therapy.

Swoger et al³⁹ treated 72 patients who had symptoms consistent with LPR with a PPI for 4 months; 25 patients in this group had less than a 50% improvement despite maximal drug therapy. Ten of these patients underwent surgical fundoplication, and 15 remained on drug therapy alone. At 1 year of follow-up, only one surgical patient (10%) reported improvement in laryngeal symptoms.

In view of this report and prior studies of surgical fundoplication,⁴⁰ surgery is not recommended for patients whose symptoms do not respond to aggressive PPI therapy.

IF A PPI FAILS, LOOK FOR OTHER CAUSES OF SYMPTOMS

Although gastroesophageal reflux and laryngeal signs and symptoms have been associated with one another, this relation may have been overstated, leading to the overdiagnosis of LPR.

The diagnosis of LPR is difficult, as laryngoscopy has high interrater variability and as the results of pH monitoring do not dependably predict who will respond to treatment.

Because PPI therapy is easy and appears to be safe, patients with extraesophageal symptoms thought to be related to reflux should undergo a trial of twice-daily PPI therapy for at least 2 months. If the patient responds to therapy, then tapering to once-daily therapy initially and then to minimal acid suppression to control symptoms would be prudent.

In patients who show no improvement, other causes of symptoms should be explored. Diseases that can mimic LPR include postnasal drip, allergies, sinus inflammation, and various pulmonary diseases. These patients should also be advised to adopt lifestyle modifications—eg, to stop smoking, lose weight, and decrease activities that cause stress on the voice. Surgery is not likely to provide any benefit in this situation. The patient should be tapered off the PPI to make sure no rebound acid reflux occurs.

The scenario is common: a patient complains of chronic hoarseness, cough, throat-clearing, sore throat, dysphagia, or a lump in the throat and undergoes laryngoscopy. If this test rules out cancer, the patient is given a diagnosis of laryngopharyngeal reflux (LPR), ie, a form of gastroesophageal reflux disease (GERD) in which the stomach contents get all the way up into the pharynx and down into the larynx. A proton pump inhibitor (PPI) is often prescribed, usually twice daily for 2 months.^1–6

In this article, we review the current understanding of the pathophysiology of LPR and evaluate current diagnostic tests and treatment regimens for patients with suspected LPR.

THE PATHOPHYSIOLOGY OF LPR IS POORLY UNDERSTOOD

Transient relaxation of the lower esophageal sphincter

In a study in 10 healthy volunteers, Dent et al⁷ found that the pressure in the lower esophageal sphincter varies considerably over a 12-hour period. Episodes of reflux were not related to low basal (resting) pressure. Rather, 70% to 100% of reflux episodes occurred during random episodes of transient, complete, and inappropriate relaxation of the sphincter that lasted about 5 to 30 seconds. The mechanism of this relaxation is not known but is thought to be related to activation of the vagus nerve, possibly as a consequence of gastric distention.⁸

Gastric, not duodenal products seem to cause the damage

In a study in dogs, Adhami et al⁹ evaluated the possible role of gastric juices (acid and pepsin) vs duodenal juices (bile acids and trypsin) in laryngeal tissue damage. After taking baseline biopsy samples of the larynx, the investigators applied a variety of gastric and duodenal enzymes at varying pH levels (pH 1–7) to the larynxes. After 9 to 12 applications, they took another biopsy and assessed the changes visually and histologically.

At low (ie, acidic) pH levels, pepsin and conjugated bile acids were the most injurious, causing erythema and histologic evidence of inflammation. The authors concluded that gastric and not duodenal substances cause laryngeal injury and that acid-suppressive therapy “should eliminate the injurious potential” of acid reflux.⁹

The larynx is more sensitive than the esophagus

Monitoring of esophageal pH has shown that healthy people can tolerate as many as 50 episodes a day of acid reflux (pH < 4) in the esophagus. However, Koufman¹⁰ found that as few as three episodes of laryngeal reflux per week can cause severe laryngeal inflammation and injury.

Does pepsin deplete buffers, worsening acid damage?

Johnston et al¹¹ took biopsies from a control group of healthy volunteers and from patients diagnosed with LPR. They detected pepsin in the samples from eight of the nine patients with LPR but in none of the controls. Furthermore, the tissue from patients with LPR had low levels of carbonic anhydrase III. The authors hypothesized that pepsin depletes the laryngopharynx of carbonic anhydrase III, and that therefore these tissues cannot produce enough bicarbonate to buffer the gastric acid. Less bicarbonate would mean greater acidity, so the pepsin would remain active and would be more likely to cause cellular damage.¹¹

However, this contention is controversial. What is universally agreed upon is that reflux of gastric or gastroduodenal contents is most likely causing injury, most likely through direct exposure, although indirect effects through vagal mechanisms cannot be ruled out.

CURRENT DIAGNOSTIC TESTS FOR LPR HAVE SHORTCOMINGS

A careful history is important. Many patients report they have sore throat, hoarseness, cough, dysphasia, or chronic throat-clearing.¹³ Factors that may predispose a patient to esophageal reflux should be discussed, eg:

• Tobacco use
• Diet (eg, soda, spicy foods, fatty foods)
• Alcohol use
• Certain drugs (calcium channel blockers, nitrates, steroids).

Up to 50% of patients presenting with extraesophageal symptoms may not have classic reflux symptoms such as heartburn and regur-gitation.¹⁴ However, the existence of “silent reflux” is currently controversial.

Because the key symptoms of LPR are nonspecific, many patients who present to an otolaryngologist undergo laryngoscopy, mainly to rule out malignancy. Once cancer is ruled out, many patients are given a diagnosis of LPR.

Furthermore, Branski et al¹⁷ performed transoral rigid laryngoscopy with videorecording in 100 consecutive patients presenting with a chief complaint of dysphonia. Five board-certified otolaryngologists individually viewed each recording, scored the degree of erythema and edema, and assessed the likelihood that LPR played a role in dysphonia and the severity of the LPR findings. The physicians’ ratings showed considerable interobserver variability. In other words, this study showed that laryngeal findings are often nonspecific and that the laryngoscopic diagnosis of LPR tends to be subjective.¹⁷

The Reflux Finding Score. Concerned by the lack of consistency in the diagnosis of LPR, Belafsky et al¹⁸ created a scoring system for documenting the physical findings and severity of disease on a standardized scale. Their Reflux Finding Score is based on eight laryngoscopic findings: subglottic edema, ventricular edema, erythema, vocal cord edema, diffuse laryngeal edema, hypertrophy of the posterior commissure, granuloma or granulation tissue, and thick endolaryngeal mucus. The total score can range from 0 (best) to 26 (worst).

In 40 patients with LPR confirmed by pH monitoring, the mean score was 11.5, compared with 5.2 in 40 age-matched controls. The authors calculated they could be 95% certain that a person with a score higher than 7 has LPR.¹⁸

However, this diagnostic method has not been validated in a large-scale randomized trial and so has yet to be incorporated into routine otolaryngology practice.

Ambulatory pH monitoring is not so golden for diagnosing LPR

Although pH monitoring was once the gold standard for diagnosing reflux, it has since been shown to be unreliable in patients who have laryngeal symptoms.⁴

How high or low in the esophagus the probe is placed is clearly critical for useful results. ⁴ But the test is subject to variability: different physicians place the probe in different locations, and the probe may shift. Another problem is that reflux may occur during untested periods.¹⁹

A pH of less than 4 in the esophagus had originally been shown to have high sensitivity and specificity,²⁰ but Reichel and Issing²¹ suggested using a pH of less than 5 as the cutoff, which would identify more patients as having LPR. Further trials are needed to more precisely determine the pH threshold for the diagnosis of LPR.

Enthusiasm is waning for pharyngeal pH monitoring

In LPR, it was initially thought that pH monitoring in the pharynx was more accurate than in the distal or proximal esophagus.

Shaker et al²² monitored the pH in the pharynx, proximal esophagus, and distal esophagus in four groups: 14 patients who had both laryngeal signs and symptoms, 12 patients who had laryngeal symptoms only, 16 patients who had GERD but no laryngeal symptoms, and 12 healthy volunteers. They found that pharyngeal reflux was more frequent and in greater quantity in patients with laryngeal signs and symptoms than in the other groups. This study suggested that pharyngeal pH monitoring may be useful in diagnosing LPR in patients who have laryngeal signs and symptoms.

However, hypopharyngeal pH monitoring has several problems. One issue is that, even in this trial, 2 of 12 healthy volunteers had episodes of pharyngeal reflux.²² In other studies, the rate of false-positive results ranged from 7% to 17%.^23,24 Additionally, in 12 previous studies, only 54% of 1,217 patients with suspected LPR had esophageal acid exposure, regardless of where the pH probe was placed.²⁵

More importantly, another study found that patients with pharyngeal reflux documented by pH monitoring were no more likely to respond to acid-suppressive therapy than patients with no documented reflux.²⁶ These findings dampen the enthusiasm for pharyngeal pH monitoring in LPR.

Impedance monitoring on therapy may be useful in refractory cases

Esophageal impedance monitoring, a newer test, uses a catheter that measures electrical resistance (impedance) between different points along the esophagus. Thus, it can detect the reflux of acid and nonacid liquid or gaseous material.

Pritchett et al²⁷ performed esophageal impedance and pH monitoring in 39 patients who were on twice-daily PPI therapy and then evaluated the same patients with wireless pH monitoring while they were off therapy. The most prevalent complaint in the study group was cough (56%), followed by heartburn (18%) and sore throat (10%).

Of the 39 patients, 25 (64%) had normal results on impedance/pH monitoring while on therapy, ruling out reflux. On pH monitoring off therapy, 28 (72%) of the 39 patients had abnormal results; this group included 13 (93%) of the 14 patients who had abnormal results on impedance/pH monitoring while on therapy. The authors recommended on-therapy testing with impedance monitoring in patients with refractory reflux, since it provides more useful clinical information.²⁷ If the results of impedance/pH monitoring are negative in these patients, a diagnosis other than reflux should be considered.

EMPIRIC PPI TREATMENT HAS SHOWN DISAPPOINTING RESULTS

Because laryngoscopy and pH monitoring are not very sensitive or specific for LPR, experts recommend empiric therapy with a PPI twice daily. However, the results have been disappointing when PPIs were compared with placebo in clinical trials.

In a randomized controlled trial,28 we found that patients who had complaints of chronic throat-clearing, cough, globus, sore throat, and hoarseness had a similar response to twice-daily esomeprazole (Nexium) compared with placebo: their primary symptom had resolved by 16 weeks in 14.7% of the esomeprazole group vs 16.0% of the placebo group (P = .799). Similarly, the final findings on laryngoscopy such as edema, erythema, and surface irregularity were not significantly different between groups.

From Qadeer MA, et al. Proton pump inhibitor therapy for suspected GERD-related chronic laryngitis: a meta analysis of randomized controlled trials. Am J Gastroenterol 2006; 101:2646–2654. Used with permission from Nature Publishing Group.

Adding a histamine-2 receptor antagonist is not recommended

Adding a histamine-2 receptor antagonist to PPI therapy has also been considered as a treatment for LPR.

Fackler et al³⁸ studied 16 GERD patients and 18 healthy volunteers to determine if adding ranitidine (Pepcid) to the PPI omeprazole (Prilosec) could improve GERD symptoms. Patients underwent baseline manometry and then gastroesophageal pH monitoring before starting the drugs. They first received omeprazole 20 mg twice daily alone for 2 weeks, and then added ranitidine 300 mg at bedtime. A pH test was done again after the first day of treatment with ranitidine, at the end of 1 week of combination therapy, and after 4 weeks of combination therapy. The combination reduced nocturnal acid breakthrough on day 1; however, due to tolerance to ranitidine, no significant difference in acid suppression was seen after 1 week of therapy. Therefore, this combination is not recommended.

Surgery is not recommended either

Some experts have argued for surgical fundoplication in patients whose symptoms persist despite drug therapy.

Swoger et al³⁹ treated 72 patients who had symptoms consistent with LPR with a PPI for 4 months; 25 patients in this group had less than a 50% improvement despite maximal drug therapy. Ten of these patients underwent surgical fundoplication, and 15 remained on drug therapy alone. At 1 year of follow-up, only one surgical patient (10%) reported improvement in laryngeal symptoms.

In view of this report and prior studies of surgical fundoplication,⁴⁰ surgery is not recommended for patients whose symptoms do not respond to aggressive PPI therapy.

IF A PPI FAILS, LOOK FOR OTHER CAUSES OF SYMPTOMS

Although gastroesophageal reflux and laryngeal signs and symptoms have been associated with one another, this relation may have been overstated, leading to the overdiagnosis of LPR.

The diagnosis of LPR is difficult, as laryngoscopy has high interrater variability and as the results of pH monitoring do not dependably predict who will respond to treatment.

Because PPI therapy is easy and appears to be safe, patients with extraesophageal symptoms thought to be related to reflux should undergo a trial of twice-daily PPI therapy for at least 2 months. If the patient responds to therapy, then tapering to once-daily therapy initially and then to minimal acid suppression to control symptoms would be prudent.

In patients who show no improvement, other causes of symptoms should be explored. Diseases that can mimic LPR include postnasal drip, allergies, sinus inflammation, and various pulmonary diseases. These patients should also be advised to adopt lifestyle modifications—eg, to stop smoking, lose weight, and decrease activities that cause stress on the voice. Surgery is not likely to provide any benefit in this situation. The patient should be tapered off the PPI to make sure no rebound acid reflux occurs.

Adequate bowel preparation depends on the right choice of bowel-cleansing agent. But with a myriad of products available, the right choice can be confusing to make.

This review discusses the currently recommended methods for bowel preparation before colonoscopy and suggests ways to solve common problems.

EARLY DETECTION IS KEY

Colorectal cancer is the third most common cancer in the United States and the second most common cause of cancer deaths. It largely can be prevented by detecting and removing adenomatous polyps, and survival rates are significantly better when it is diagnosed while still localized.⁵ Early detection, through widely applied screening programs that include colonoscopy, is thought to be playing a key role in the recent decline of colorectal cancer rates in developed countries.⁶

THREE TYPES OF AGENTS

Bowel preparation agents, for the most part, can be classified into one of three categories:

• Polyethylene glycol solutions, which work as high-volume gut lavage solutions
• Osmotic agents, such as sodium phosphate, magnesium citrate, lactulose, and mannitol, which draw extracellular fluid across the bowel wall and into the lumen
• Stimulants (castor oil, senna, sodium picosulfte, and bisacodyl), which work by increasing smooth muscle activity within the wall of the colon.

POLYETHYLENE GLYCOL SOLUTIONS

Bowel preparation in the past consisted of dietary restriction, stimulant laxatives, and enemas. ^7,8 However, these were time-consuming (taking 48–72 hours), harsh, and not very effective for adequate visualization during colonoscopy.

In 1980, Davis et al⁹ developed an osmotically balanced, high-molecular weight, nonabsorbable polymer given in a dilute electrolyte solution. The osmotic effect of the polymer keeps the electrolyte solution in the colon. Since little fluid is exchanged across the colonic membrane, the potential for systemic electrolyte disturbance is limited.

Since then, these solutions have become some of the preferred bowel cleansing agents worldwide.^7,8 They work as an oral lavage and hence need to be taken in high volume (typically 4 L) for bowel cleansing.

Advantages and disadvantages of polyethylene glycol solutions

Polyethylene glycol solutions are more effective and better tolerated than regimens of diet combined with cathartic agents, or high-volume balanced electrolyte solutions, or mannitol-based solutions.⁷ Since they are osmotically balanced and do not induce substantial shifts in fluid and electrolytes, they are safe for patients who have electrolyte imbalances, advanced liver disease, poorly compensated congestive heart failure, or renal failure.

These preparations are, however, contraindicated in patients who have allergies to polyethylene glycol compounds, gastric outlet obstruction, high-grade small-bowel obstruction, significant colonic obstruction, perforation, diverticulitis, or hemodynamic instability. In addition, they are classified by the US Food and Drug Administration (FDA) as pregnancy category C and have been associated (albeit rarely) with Mallory-Weiss tear, toxic colitis, pulmonary aspiration, hypothermia, cardiac arrhythmias, pancreatitis, and inappropriate antidiuretic hormone secretion.^10,11

The main disadvantages of these solutions are the large volume of fluid (4 L) that patients must drink and their unpalatable taste, which is due to sodium sulfate. The large volume of ingestion is the main reason for nausea, bloating, cramping, and vomiting with these products, which affect patient compliance and the ultimate success of colonoscopy.

Commercially available polyethylene glycol solutions

Standard full-volume solutions (Colyte, GoLYTELY) have been widely studied and have the most evidence of safety and effectiveness. They are also inexpensive, and most insurance companies pay for them. However, about 5% to 15% of patients do not complete the preparation, because of poor palatability, large volume, or both.⁷

Sulfate-free and flavored solutions. To make polyethylene glycol solutions more tolerable, sulfate-free solutions have been developed. These are less salty, more palatable, and comparable to standard solutions in terms of effective colonic cleansing.¹² Sulfate-free polyethylene glycol solutions commercially available in the United States are NuLytely (flavors: cherry, lemon-lime, orange, pineapple) and TriLyte (flavors: cherry, citrus-berry, lemon-lime, orange, pineapple).

Low-volume solutions have been developed in an attempt to increase acceptability and reduce volume-related adverse effects such as bloating. For example, HalfLytely (flavor: lemon-lime) consists of 2 L of polyethylene glycol solution packaged with two bisacodyl tablets. Stimulant laxatives such as bisacodyl and magnesium citrate effectively debulk the colon of solid stool and allow a lower volume of solution to be used.^13,14

Also commercially available is a preparation that contains ascorbic acid (MoviPrep). Ascorbic acid acts as a flavoring and as a cathartic, also permitting a lower volume of fluid to be used.

Studies that compared full-volume and low-volume regimens (the latter including ascorbic acid, magnesium citrate, or bisacodyl) found the low-volume regimens to be as effective and more tolerable.^14–18

Combining over-the-counter polyethylene glycol 3350 laxative powder (MiraLAX) and Gatorade or Crystal Light (or another clear liquid of choice) has also been shown to improve the taste and tolerability of the preparation. Although beneficial and commonly used in certain regions of the United States, this combination is not approved for bowel preparation and its use is considered off-label.

One way to increase tolerability and patient adherence is to split the dose so that the patient takes half the laxative prescription (polyethylene glycol or otherwise) the night before colonoscopy and the other half in the morning, usually about 4 to 5 hours before the scheduled time of the procedure.^18,19

Split dosing not only improves patient acceptability, but also cleans the colon better.⁴ Traditional dosing, ie, drinking the entire volume of solution the night before, leaves a long interval between the end of the preparation process and the start of the procedure. Thick intestinal secretions empty out of the small intestine during that interval and obscure the cecum and ascending colon. With split dosing, the second dose is completed a few hours before the procedure, cleaning out the remaining intestinal secretions and obviating this problem.

Other measures that can make polyethylene glycol solutions more tolerable are:

• Chilling the solution
• Adding lemon slices or sugar-free flavor enhancers (such as Crystal Light) or lemon juice
• Giving the solution by nasogastric tube (at a rate of 1.2–1.8 L per hour) in patients with swallowing dysfunction or altered mental status
• Adding metoclopramide (Reglan) 5 to 10 mg orally to prevent or treat nausea
• Adding magnesium citrate (1 bottle, about 300 mL) in patients without renal insufficiency, or bisacodyl (two to four tablets of 5 mg each), so that the volume can be less^15,16
• Stopping further ingestion of solution once the stool is watery and clear on the morning of the procedure (for patients who can clearly understand and follow bowel preparation instructions).¹⁷

SODIUM PHOSPHATE SOLUTIONS

Sodium phosphate is an osmotic laxative that draws water into the bowel lumen to promote colonic cleansing. Retention of water in the lumen of the colon stimulates peristalsis and bowel movements.

Advantages and disadvantages of sodium phosphate solutions

Sodium phosphate is widely used worldwide and has been found to be a very acceptable and effective bowel cleansing agent. A recent systematic review of 25 studies¹⁸ found that sodium phosphate was superior to polyethylene glycol in 14 studies, that there was no significant difference in 10 studies, and that only one study found polyethylene glycol to be better tolerated than sodium phosphate.¹⁸ Similarly, a meta-analysis¹⁹ found sodium phosphate to be more effective than polyethylene glycol in bowel cleansing (odds ratio 0.75; P = .0004); more easily completed by patients (odds ratio 0.16; P < .00001); and comparable in terms of adverse events (odds ratio 0.98; P = .81).¹⁹ However, most of the clinical trials excluded patients who had renal failure, ascites, or serious heart disease—the groups most at risk of significant adverse effects from sodium phosphate use. The main reasons sodium phosphate was better tolerated were better flavor and smaller volume (1.5–2 L compared with 4 L for polyethylene glycol).^20–22

The main disadvantage of sodium phosphate is its potential to cause large fluid and electrolyte shifts. Its use has been associated with a variety of electrolyte abnormalities, including hyperphosphatemia, hypocalcemia, hypokalemia, increased plasma osmolality, hyponatremia, and, conversely, hypernatremia.^7,8,23 Asymptomatic hyperphosphatemia alone can be seen in as many as 40% of healthy patients completing sodium phosphate preparations. It may be significant in patients with renal failure and can lead to acute phosphate nephropathy.

Rare adverse events such as nephrocalcinosis with acute renal failure also have been reported, especially in patients taking angiotensin-converting enzyme (ACE) inhibitors or angiotensin receptor blockers.²³

The significant volume contraction and consequent dehydration seen in some patients using sodium phosphate may be decreased by encouraging patients to drink fluids liberally, especially before the day of the procedure and after the procedure.⁷

Recently, renal failure due to hyperphosphatemia (acute phosphate nephropathy) has been reported even in patients with normal kidney function.²⁴ Because of the risk of inappropriate use or overdose associated with over-the counter sodium phosphate, the FDA recommended on December 11, 2008, that sodium phosphate products be available only by prescription when they are used for bowel cleansing.²⁵ The C.B. Fleet Company voluntarily recalled its oral sodium phosphate products sold over the counter (Fleet Phospho-Soda and Fleet EZ-PREP). In addition, the FDA required a black box warning on the prescription oral sodium phosphate products Visicol and OsmoPrep, alerting consumers to the risk of acute phosphate nephropathy.²⁵ According to the FDA, health professionals should use caution when prescribing Visicol or OsmoPrep for patients who may be at higher risk of kidney injury, such as:

• Patients over 55 years of age
• Patients who are dehydrated or who have kidney disease, acute colitis, or delayed bowel emptying
• Patients taking certain drugs that affect kidney function, such as diuretics, ACE inhibitors, angiotensin receptor blockers, and nonsteroidal anti-inflammatory drugs.¹⁶

Commercially available sodium phosphate products

Sodium phosphate products can still be prescribed, but they are no longer available over the counter in the United States. Patients should be screened to make sure they can safely take these products, and the doses should not exceed the maximum recommended.

Figure 2 shows a simplified algorithm for selecting the optimal bowel preparation agent for an individual patient.

OTHER BOWEL PREPARATION AGENTS AND ADJUNCTS

Magnesium citrate

Like sodium phosphate, magnesium citrate is a hyperosmotic agent that promotes bowel cleansing by increasing intraluminal fluid volume. Since magnesium is eliminated solely by the kidney, it should be used with extreme caution in patients with renal insufficiency or renal failure.

Adding magnesium citrate as an adjunct to polyethylene glycol has been shown to reduce the amount of polyethylene glycol solution required (2 L) for the same result.¹⁷

For patients who cannot tolerate polyethylene glycol, a reasonable alternative is magnesium citrate (1 bottle, around 300 mL) the evening before the procedure plus either bisacodyl tablets at the same time as the magnesium citrate or rectal pulsed irrigation immediately before the procedure.⁷

Saline laxatives that include sodium picosulfate and magnesium citrate in combination are available primarily in the United Kingdom for bowel preparation for colonoscopy. Sodium picosulfate acts locally in the colon as a stimulant laxative and by increasing the force of laxatives, whereas magnesium citrate acts as an osmotic laxative by retaining fluids in the colon to clear the colon and rectum of fecal contents. The combination has been found to have similar efficacy and tolerability as sodium phosphate but is not currently available in the United States.²⁶

Enemas

Enemas are sufficient for flexible sigmoidoscopy, but when used alone they do not clean out the proximal colon enough for adequate visualization during colonoscopy. They are best used as adjuncts to other bowel preparation agents when patients present with poor distal colon preparation for colonoscopy.^7,27 Enemas are also useful in washing out the distal segment of bowel in patients with a proximal stoma. The common types of enemas used are tap water, sodium biphosphate (Fleet), and mineral oil.

Tap water enemas distend the rectum and mimic the natural distention by the stool to allow the rectum to empty itself. Tap water (1 L) has fewer adverse effects than sodium biphosphate or mineral oil but is less effective.

Sodium biphosphate (Fleet) enemas draw fluid into the bowel by osmotic action, prompting contraction. One or two bottles are commonly used for bowel cleansing before sigmoidoscopy. However, as with oral sodium phosphate, sodium biphosphate enemas should be avoided in the elderly and in those with renal failure because of the risk of hyperphosphatemia and subsequent hypocalcemia.

In a head-to-head comparison,²⁸ sodium biphosphate enema was found to provide significantly better bowel preparation than the sodium picosulfate-magnesium citrate combination (currently not available in the United States) for flexible sigmoidoscopy, being judged adequate or better in 93% of procedures as opposed to 74%.²⁸

Oil-based enemas such as cottonseed oil plus docusate (Colace) and diatrizoate sodium (Hypaque) are powerful lubricant laxatives that work by slowing the absorption of water from the bowel, so that the stool is softer. However, they have a number of adverse effects, such as severe allergic reactions (including angioedema and anaphylaxis), muscle cramps, and sporadic seepage that can soil the patient's undergarments for up to 24 hours. Also, their safety in children less than 2 years of age and in pregnant and breastfeeding mothers is not established.

Oil-based enemas are usually reserved for short-term use in refractory constipation, especially to soften feces that has become hardened within the rectum (as in fecal impaction).²⁷

Adjuncts

Diet. Dietary modifications alone, such as a clear liquid diet, are inadequate for colonoscopy, but they may be beneficial as adjuncts to other cleansing methods by decreasing the formation of solid residue. Clear liquids also help maintain adequate hydration during bowel preparation and are recommended with all bowel preparation regimens.

Hyperosmolar or stimulant laxatives. Bisacodyl (two to four tablets of 5 mg each), magnesium citrate (one bottle, about 300 mL), and low-dose senna (36 mg, about four 8.6-mg Sennakot tablets) have been used as adjuncts to low-volume polyethylene glycol solution, achieving results similar to those with full-volume polyethylene glycol. Depending on the type of study to be done, these agents are taken within 2 to 6 hours of starting the polyethylene glycol solution.

In contrast, the routine use of nonabsorbable carbohydrates such as mannitol and lactulose is not favored for bowel preparation, since the hydrogen gas produced by bacterial fermentation of the nonabsorbed carbohydrates increases the risk of explosion during electrosurgical procedures.²⁹

Antiemetic agents. Metoclopramide (5–10 mg), a dopamine antagonist gastroprokinetic that sensitizes tissues to the action of acetylcholine, is commonly used to prevent nausea or vomiting associated with bowel preparation agents.^7,30

Antifoaming agent. Simethicone (three tablets of 80 mg each, total dose 240 mg), an anti-flatulent, anti-gas agent, is prescribed by many gastroenterologists in an attempt to reduce bubbles during colonoscopy and improve visibility. It works by reducing the surface tension of air bubbles and causing small bubbles to coalesce into larger ones that pass more easily with belching or flatulence.

Nasogastric or orogastric tubes have been used to instill colonic preparations, especially for inpatients unable to drink polyethylene glycol solutions or for patients who are unresponsive or mechanically ventilated. This method can also be useful for rapid bowel cleansing (within 2 to 3 hours) for patients with lower gastrointestinal bleeding. However, routine use of a nasogastric tube solely for bowel preparation is discouraged as it can lead to severe complications, such as aspiration and trauma during insertion.⁷

OTHER CONSIDERATIONS

Patient education

The importance of patient education for successful bowel preparation cannot be overemphasized. Patients need to be informed about why they need to undergo colonoscopy, the importance of bowel preparation, the side effects of agents used, and the exact preparation instructions. An interactive educational tutorial about colonoscopy for patients is available at Medline Plus at http://www.nlm.nih.gov/medlineplus/tutorials/colonoscopy/htm/index.htm.

Role of hydration

A commonly held misconception is that patients taking 4 L of polyethylene glycol do not need additional hydration, since they are already ingesting such a large volume of fluid. Given that bowel preparations induce diarrhea and, in some instances, nausea and vomiting, all patients taking bowel preparations are at risk of dehydration.³² In fact, the fluid loss during bowel preparation may exceed 2 to 3 L. It is not surprising that many safety issues associated with bowel preparation agents are related to dehydration and its complications.

Hence, patients should be advised to consume at least 64 oz (approximately 2 L) of clear fluid on the day before the colonoscopy. According to the American Society of Anesthesiologists, clear liquids can be safely ingested up until 2 hours before receiving anesthesia.³³ Patients should contact their physicians if they experience vomiting or cannot comply with clear liquid volume instructions prior to colonoscopy. Metoclopramide has been found useful in many cases of nausea or vomiting associated with bowel preparation agents.¹⁸ In addition, patients should also be reminded to keep drinking extra fluids after the procedure is completed to reduce the risk of dehydration and its complications (Table 2).

Adequate bowel preparation depends on the right choice of bowel-cleansing agent. But with a myriad of products available, the right choice can be confusing to make.

This review discusses the currently recommended methods for bowel preparation before colonoscopy and suggests ways to solve common problems.

EARLY DETECTION IS KEY

Colorectal cancer is the third most common cancer in the United States and the second most common cause of cancer deaths. It largely can be prevented by detecting and removing adenomatous polyps, and survival rates are significantly better when it is diagnosed while still localized.⁵ Early detection, through widely applied screening programs that include colonoscopy, is thought to be playing a key role in the recent decline of colorectal cancer rates in developed countries.⁶

THREE TYPES OF AGENTS

Bowel preparation agents, for the most part, can be classified into one of three categories:

• Polyethylene glycol solutions, which work as high-volume gut lavage solutions
• Osmotic agents, such as sodium phosphate, magnesium citrate, lactulose, and mannitol, which draw extracellular fluid across the bowel wall and into the lumen
• Stimulants (castor oil, senna, sodium picosulfte, and bisacodyl), which work by increasing smooth muscle activity within the wall of the colon.

POLYETHYLENE GLYCOL SOLUTIONS

Bowel preparation in the past consisted of dietary restriction, stimulant laxatives, and enemas. ^7,8 However, these were time-consuming (taking 48–72 hours), harsh, and not very effective for adequate visualization during colonoscopy.

In 1980, Davis et al⁹ developed an osmotically balanced, high-molecular weight, nonabsorbable polymer given in a dilute electrolyte solution. The osmotic effect of the polymer keeps the electrolyte solution in the colon. Since little fluid is exchanged across the colonic membrane, the potential for systemic electrolyte disturbance is limited.

Since then, these solutions have become some of the preferred bowel cleansing agents worldwide.^7,8 They work as an oral lavage and hence need to be taken in high volume (typically 4 L) for bowel cleansing.

Advantages and disadvantages of polyethylene glycol solutions

Polyethylene glycol solutions are more effective and better tolerated than regimens of diet combined with cathartic agents, or high-volume balanced electrolyte solutions, or mannitol-based solutions.⁷ Since they are osmotically balanced and do not induce substantial shifts in fluid and electrolytes, they are safe for patients who have electrolyte imbalances, advanced liver disease, poorly compensated congestive heart failure, or renal failure.

These preparations are, however, contraindicated in patients who have allergies to polyethylene glycol compounds, gastric outlet obstruction, high-grade small-bowel obstruction, significant colonic obstruction, perforation, diverticulitis, or hemodynamic instability. In addition, they are classified by the US Food and Drug Administration (FDA) as pregnancy category C and have been associated (albeit rarely) with Mallory-Weiss tear, toxic colitis, pulmonary aspiration, hypothermia, cardiac arrhythmias, pancreatitis, and inappropriate antidiuretic hormone secretion.^10,11

The main disadvantages of these solutions are the large volume of fluid (4 L) that patients must drink and their unpalatable taste, which is due to sodium sulfate. The large volume of ingestion is the main reason for nausea, bloating, cramping, and vomiting with these products, which affect patient compliance and the ultimate success of colonoscopy.

Commercially available polyethylene glycol solutions

Standard full-volume solutions (Colyte, GoLYTELY) have been widely studied and have the most evidence of safety and effectiveness. They are also inexpensive, and most insurance companies pay for them. However, about 5% to 15% of patients do not complete the preparation, because of poor palatability, large volume, or both.⁷

Sulfate-free and flavored solutions. To make polyethylene glycol solutions more tolerable, sulfate-free solutions have been developed. These are less salty, more palatable, and comparable to standard solutions in terms of effective colonic cleansing.¹² Sulfate-free polyethylene glycol solutions commercially available in the United States are NuLytely (flavors: cherry, lemon-lime, orange, pineapple) and TriLyte (flavors: cherry, citrus-berry, lemon-lime, orange, pineapple).

Low-volume solutions have been developed in an attempt to increase acceptability and reduce volume-related adverse effects such as bloating. For example, HalfLytely (flavor: lemon-lime) consists of 2 L of polyethylene glycol solution packaged with two bisacodyl tablets. Stimulant laxatives such as bisacodyl and magnesium citrate effectively debulk the colon of solid stool and allow a lower volume of solution to be used.^13,14

Also commercially available is a preparation that contains ascorbic acid (MoviPrep). Ascorbic acid acts as a flavoring and as a cathartic, also permitting a lower volume of fluid to be used.

Studies that compared full-volume and low-volume regimens (the latter including ascorbic acid, magnesium citrate, or bisacodyl) found the low-volume regimens to be as effective and more tolerable.^14–18

Combining over-the-counter polyethylene glycol 3350 laxative powder (MiraLAX) and Gatorade or Crystal Light (or another clear liquid of choice) has also been shown to improve the taste and tolerability of the preparation. Although beneficial and commonly used in certain regions of the United States, this combination is not approved for bowel preparation and its use is considered off-label.

One way to increase tolerability and patient adherence is to split the dose so that the patient takes half the laxative prescription (polyethylene glycol or otherwise) the night before colonoscopy and the other half in the morning, usually about 4 to 5 hours before the scheduled time of the procedure.^18,19

Split dosing not only improves patient acceptability, but also cleans the colon better.⁴ Traditional dosing, ie, drinking the entire volume of solution the night before, leaves a long interval between the end of the preparation process and the start of the procedure. Thick intestinal secretions empty out of the small intestine during that interval and obscure the cecum and ascending colon. With split dosing, the second dose is completed a few hours before the procedure, cleaning out the remaining intestinal secretions and obviating this problem.

Other measures that can make polyethylene glycol solutions more tolerable are:

• Chilling the solution
• Adding lemon slices or sugar-free flavor enhancers (such as Crystal Light) or lemon juice
• Giving the solution by nasogastric tube (at a rate of 1.2–1.8 L per hour) in patients with swallowing dysfunction or altered mental status
• Adding metoclopramide (Reglan) 5 to 10 mg orally to prevent or treat nausea
• Adding magnesium citrate (1 bottle, about 300 mL) in patients without renal insufficiency, or bisacodyl (two to four tablets of 5 mg each), so that the volume can be less^15,16
• Stopping further ingestion of solution once the stool is watery and clear on the morning of the procedure (for patients who can clearly understand and follow bowel preparation instructions).¹⁷

SODIUM PHOSPHATE SOLUTIONS

Sodium phosphate is an osmotic laxative that draws water into the bowel lumen to promote colonic cleansing. Retention of water in the lumen of the colon stimulates peristalsis and bowel movements.

Advantages and disadvantages of sodium phosphate solutions

Sodium phosphate is widely used worldwide and has been found to be a very acceptable and effective bowel cleansing agent. A recent systematic review of 25 studies¹⁸ found that sodium phosphate was superior to polyethylene glycol in 14 studies, that there was no significant difference in 10 studies, and that only one study found polyethylene glycol to be better tolerated than sodium phosphate.¹⁸ Similarly, a meta-analysis¹⁹ found sodium phosphate to be more effective than polyethylene glycol in bowel cleansing (odds ratio 0.75; P = .0004); more easily completed by patients (odds ratio 0.16; P < .00001); and comparable in terms of adverse events (odds ratio 0.98; P = .81).¹⁹ However, most of the clinical trials excluded patients who had renal failure, ascites, or serious heart disease—the groups most at risk of significant adverse effects from sodium phosphate use. The main reasons sodium phosphate was better tolerated were better flavor and smaller volume (1.5–2 L compared with 4 L for polyethylene glycol).^20–22

The main disadvantage of sodium phosphate is its potential to cause large fluid and electrolyte shifts. Its use has been associated with a variety of electrolyte abnormalities, including hyperphosphatemia, hypocalcemia, hypokalemia, increased plasma osmolality, hyponatremia, and, conversely, hypernatremia.^7,8,23 Asymptomatic hyperphosphatemia alone can be seen in as many as 40% of healthy patients completing sodium phosphate preparations. It may be significant in patients with renal failure and can lead to acute phosphate nephropathy.

Rare adverse events such as nephrocalcinosis with acute renal failure also have been reported, especially in patients taking angiotensin-converting enzyme (ACE) inhibitors or angiotensin receptor blockers.²³

The significant volume contraction and consequent dehydration seen in some patients using sodium phosphate may be decreased by encouraging patients to drink fluids liberally, especially before the day of the procedure and after the procedure.⁷

Recently, renal failure due to hyperphosphatemia (acute phosphate nephropathy) has been reported even in patients with normal kidney function.²⁴ Because of the risk of inappropriate use or overdose associated with over-the counter sodium phosphate, the FDA recommended on December 11, 2008, that sodium phosphate products be available only by prescription when they are used for bowel cleansing.²⁵ The C.B. Fleet Company voluntarily recalled its oral sodium phosphate products sold over the counter (Fleet Phospho-Soda and Fleet EZ-PREP). In addition, the FDA required a black box warning on the prescription oral sodium phosphate products Visicol and OsmoPrep, alerting consumers to the risk of acute phosphate nephropathy.²⁵ According to the FDA, health professionals should use caution when prescribing Visicol or OsmoPrep for patients who may be at higher risk of kidney injury, such as:

• Patients over 55 years of age
• Patients who are dehydrated or who have kidney disease, acute colitis, or delayed bowel emptying
• Patients taking certain drugs that affect kidney function, such as diuretics, ACE inhibitors, angiotensin receptor blockers, and nonsteroidal anti-inflammatory drugs.¹⁶

Commercially available sodium phosphate products

Sodium phosphate products can still be prescribed, but they are no longer available over the counter in the United States. Patients should be screened to make sure they can safely take these products, and the doses should not exceed the maximum recommended.

Figure 2 shows a simplified algorithm for selecting the optimal bowel preparation agent for an individual patient.

OTHER BOWEL PREPARATION AGENTS AND ADJUNCTS

Magnesium citrate

Like sodium phosphate, magnesium citrate is a hyperosmotic agent that promotes bowel cleansing by increasing intraluminal fluid volume. Since magnesium is eliminated solely by the kidney, it should be used with extreme caution in patients with renal insufficiency or renal failure.

Adding magnesium citrate as an adjunct to polyethylene glycol has been shown to reduce the amount of polyethylene glycol solution required (2 L) for the same result.¹⁷

For patients who cannot tolerate polyethylene glycol, a reasonable alternative is magnesium citrate (1 bottle, around 300 mL) the evening before the procedure plus either bisacodyl tablets at the same time as the magnesium citrate or rectal pulsed irrigation immediately before the procedure.⁷

Saline laxatives that include sodium picosulfate and magnesium citrate in combination are available primarily in the United Kingdom for bowel preparation for colonoscopy. Sodium picosulfate acts locally in the colon as a stimulant laxative and by increasing the force of laxatives, whereas magnesium citrate acts as an osmotic laxative by retaining fluids in the colon to clear the colon and rectum of fecal contents. The combination has been found to have similar efficacy and tolerability as sodium phosphate but is not currently available in the United States.²⁶

Enemas

Enemas are sufficient for flexible sigmoidoscopy, but when used alone they do not clean out the proximal colon enough for adequate visualization during colonoscopy. They are best used as adjuncts to other bowel preparation agents when patients present with poor distal colon preparation for colonoscopy.^7,27 Enemas are also useful in washing out the distal segment of bowel in patients with a proximal stoma. The common types of enemas used are tap water, sodium biphosphate (Fleet), and mineral oil.

Tap water enemas distend the rectum and mimic the natural distention by the stool to allow the rectum to empty itself. Tap water (1 L) has fewer adverse effects than sodium biphosphate or mineral oil but is less effective.

Sodium biphosphate (Fleet) enemas draw fluid into the bowel by osmotic action, prompting contraction. One or two bottles are commonly used for bowel cleansing before sigmoidoscopy. However, as with oral sodium phosphate, sodium biphosphate enemas should be avoided in the elderly and in those with renal failure because of the risk of hyperphosphatemia and subsequent hypocalcemia.

In a head-to-head comparison,²⁸ sodium biphosphate enema was found to provide significantly better bowel preparation than the sodium picosulfate-magnesium citrate combination (currently not available in the United States) for flexible sigmoidoscopy, being judged adequate or better in 93% of procedures as opposed to 74%.²⁸

Oil-based enemas such as cottonseed oil plus docusate (Colace) and diatrizoate sodium (Hypaque) are powerful lubricant laxatives that work by slowing the absorption of water from the bowel, so that the stool is softer. However, they have a number of adverse effects, such as severe allergic reactions (including angioedema and anaphylaxis), muscle cramps, and sporadic seepage that can soil the patient's undergarments for up to 24 hours. Also, their safety in children less than 2 years of age and in pregnant and breastfeeding mothers is not established.

Oil-based enemas are usually reserved for short-term use in refractory constipation, especially to soften feces that has become hardened within the rectum (as in fecal impaction).²⁷

Adjuncts

Diet. Dietary modifications alone, such as a clear liquid diet, are inadequate for colonoscopy, but they may be beneficial as adjuncts to other cleansing methods by decreasing the formation of solid residue. Clear liquids also help maintain adequate hydration during bowel preparation and are recommended with all bowel preparation regimens.

Hyperosmolar or stimulant laxatives. Bisacodyl (two to four tablets of 5 mg each), magnesium citrate (one bottle, about 300 mL), and low-dose senna (36 mg, about four 8.6-mg Sennakot tablets) have been used as adjuncts to low-volume polyethylene glycol solution, achieving results similar to those with full-volume polyethylene glycol. Depending on the type of study to be done, these agents are taken within 2 to 6 hours of starting the polyethylene glycol solution.

In contrast, the routine use of nonabsorbable carbohydrates such as mannitol and lactulose is not favored for bowel preparation, since the hydrogen gas produced by bacterial fermentation of the nonabsorbed carbohydrates increases the risk of explosion during electrosurgical procedures.²⁹

Antiemetic agents. Metoclopramide (5–10 mg), a dopamine antagonist gastroprokinetic that sensitizes tissues to the action of acetylcholine, is commonly used to prevent nausea or vomiting associated with bowel preparation agents.^7,30

Antifoaming agent. Simethicone (three tablets of 80 mg each, total dose 240 mg), an anti-flatulent, anti-gas agent, is prescribed by many gastroenterologists in an attempt to reduce bubbles during colonoscopy and improve visibility. It works by reducing the surface tension of air bubbles and causing small bubbles to coalesce into larger ones that pass more easily with belching or flatulence.

Nasogastric or orogastric tubes have been used to instill colonic preparations, especially for inpatients unable to drink polyethylene glycol solutions or for patients who are unresponsive or mechanically ventilated. This method can also be useful for rapid bowel cleansing (within 2 to 3 hours) for patients with lower gastrointestinal bleeding. However, routine use of a nasogastric tube solely for bowel preparation is discouraged as it can lead to severe complications, such as aspiration and trauma during insertion.⁷

OTHER CONSIDERATIONS

Patient education

The importance of patient education for successful bowel preparation cannot be overemphasized. Patients need to be informed about why they need to undergo colonoscopy, the importance of bowel preparation, the side effects of agents used, and the exact preparation instructions. An interactive educational tutorial about colonoscopy for patients is available at Medline Plus at http://www.nlm.nih.gov/medlineplus/tutorials/colonoscopy/htm/index.htm.

Role of hydration

A commonly held misconception is that patients taking 4 L of polyethylene glycol do not need additional hydration, since they are already ingesting such a large volume of fluid. Given that bowel preparations induce diarrhea and, in some instances, nausea and vomiting, all patients taking bowel preparations are at risk of dehydration.³² In fact, the fluid loss during bowel preparation may exceed 2 to 3 L. It is not surprising that many safety issues associated with bowel preparation agents are related to dehydration and its complications.

Hence, patients should be advised to consume at least 64 oz (approximately 2 L) of clear fluid on the day before the colonoscopy. According to the American Society of Anesthesiologists, clear liquids can be safely ingested up until 2 hours before receiving anesthesia.³³ Patients should contact their physicians if they experience vomiting or cannot comply with clear liquid volume instructions prior to colonoscopy. Metoclopramide has been found useful in many cases of nausea or vomiting associated with bowel preparation agents.¹⁸ In addition, patients should also be reminded to keep drinking extra fluids after the procedure is completed to reduce the risk of dehydration and its complications (Table 2).

Adequate bowel preparation depends on the right choice of bowel-cleansing agent. But with a myriad of products available, the right choice can be confusing to make.

This review discusses the currently recommended methods for bowel preparation before colonoscopy and suggests ways to solve common problems.

EARLY DETECTION IS KEY

Colorectal cancer is the third most common cancer in the United States and the second most common cause of cancer deaths. It largely can be prevented by detecting and removing adenomatous polyps, and survival rates are significantly better when it is diagnosed while still localized.⁵ Early detection, through widely applied screening programs that include colonoscopy, is thought to be playing a key role in the recent decline of colorectal cancer rates in developed countries.⁶

THREE TYPES OF AGENTS

Bowel preparation agents, for the most part, can be classified into one of three categories:

• Polyethylene glycol solutions, which work as high-volume gut lavage solutions
• Osmotic agents, such as sodium phosphate, magnesium citrate, lactulose, and mannitol, which draw extracellular fluid across the bowel wall and into the lumen
• Stimulants (castor oil, senna, sodium picosulfte, and bisacodyl), which work by increasing smooth muscle activity within the wall of the colon.

POLYETHYLENE GLYCOL SOLUTIONS

Bowel preparation in the past consisted of dietary restriction, stimulant laxatives, and enemas. ^7,8 However, these were time-consuming (taking 48–72 hours), harsh, and not very effective for adequate visualization during colonoscopy.

In 1980, Davis et al⁹ developed an osmotically balanced, high-molecular weight, nonabsorbable polymer given in a dilute electrolyte solution. The osmotic effect of the polymer keeps the electrolyte solution in the colon. Since little fluid is exchanged across the colonic membrane, the potential for systemic electrolyte disturbance is limited.

Since then, these solutions have become some of the preferred bowel cleansing agents worldwide.^7,8 They work as an oral lavage and hence need to be taken in high volume (typically 4 L) for bowel cleansing.

Advantages and disadvantages of polyethylene glycol solutions

Polyethylene glycol solutions are more effective and better tolerated than regimens of diet combined with cathartic agents, or high-volume balanced electrolyte solutions, or mannitol-based solutions.⁷ Since they are osmotically balanced and do not induce substantial shifts in fluid and electrolytes, they are safe for patients who have electrolyte imbalances, advanced liver disease, poorly compensated congestive heart failure, or renal failure.

These preparations are, however, contraindicated in patients who have allergies to polyethylene glycol compounds, gastric outlet obstruction, high-grade small-bowel obstruction, significant colonic obstruction, perforation, diverticulitis, or hemodynamic instability. In addition, they are classified by the US Food and Drug Administration (FDA) as pregnancy category C and have been associated (albeit rarely) with Mallory-Weiss tear, toxic colitis, pulmonary aspiration, hypothermia, cardiac arrhythmias, pancreatitis, and inappropriate antidiuretic hormone secretion.^10,11

The main disadvantages of these solutions are the large volume of fluid (4 L) that patients must drink and their unpalatable taste, which is due to sodium sulfate. The large volume of ingestion is the main reason for nausea, bloating, cramping, and vomiting with these products, which affect patient compliance and the ultimate success of colonoscopy.

Commercially available polyethylene glycol solutions

Standard full-volume solutions (Colyte, GoLYTELY) have been widely studied and have the most evidence of safety and effectiveness. They are also inexpensive, and most insurance companies pay for them. However, about 5% to 15% of patients do not complete the preparation, because of poor palatability, large volume, or both.⁷

Sulfate-free and flavored solutions. To make polyethylene glycol solutions more tolerable, sulfate-free solutions have been developed. These are less salty, more palatable, and comparable to standard solutions in terms of effective colonic cleansing.¹² Sulfate-free polyethylene glycol solutions commercially available in the United States are NuLytely (flavors: cherry, lemon-lime, orange, pineapple) and TriLyte (flavors: cherry, citrus-berry, lemon-lime, orange, pineapple).

Low-volume solutions have been developed in an attempt to increase acceptability and reduce volume-related adverse effects such as bloating. For example, HalfLytely (flavor: lemon-lime) consists of 2 L of polyethylene glycol solution packaged with two bisacodyl tablets. Stimulant laxatives such as bisacodyl and magnesium citrate effectively debulk the colon of solid stool and allow a lower volume of solution to be used.^13,14

Also commercially available is a preparation that contains ascorbic acid (MoviPrep). Ascorbic acid acts as a flavoring and as a cathartic, also permitting a lower volume of fluid to be used.

Studies that compared full-volume and low-volume regimens (the latter including ascorbic acid, magnesium citrate, or bisacodyl) found the low-volume regimens to be as effective and more tolerable.^14–18

Combining over-the-counter polyethylene glycol 3350 laxative powder (MiraLAX) and Gatorade or Crystal Light (or another clear liquid of choice) has also been shown to improve the taste and tolerability of the preparation. Although beneficial and commonly used in certain regions of the United States, this combination is not approved for bowel preparation and its use is considered off-label.

One way to increase tolerability and patient adherence is to split the dose so that the patient takes half the laxative prescription (polyethylene glycol or otherwise) the night before colonoscopy and the other half in the morning, usually about 4 to 5 hours before the scheduled time of the procedure.^18,19

Split dosing not only improves patient acceptability, but also cleans the colon better.⁴ Traditional dosing, ie, drinking the entire volume of solution the night before, leaves a long interval between the end of the preparation process and the start of the procedure. Thick intestinal secretions empty out of the small intestine during that interval and obscure the cecum and ascending colon. With split dosing, the second dose is completed a few hours before the procedure, cleaning out the remaining intestinal secretions and obviating this problem.

Other measures that can make polyethylene glycol solutions more tolerable are:

• Chilling the solution
• Adding lemon slices or sugar-free flavor enhancers (such as Crystal Light) or lemon juice
• Giving the solution by nasogastric tube (at a rate of 1.2–1.8 L per hour) in patients with swallowing dysfunction or altered mental status
• Adding metoclopramide (Reglan) 5 to 10 mg orally to prevent or treat nausea
• Adding magnesium citrate (1 bottle, about 300 mL) in patients without renal insufficiency, or bisacodyl (two to four tablets of 5 mg each), so that the volume can be less^15,16
• Stopping further ingestion of solution once the stool is watery and clear on the morning of the procedure (for patients who can clearly understand and follow bowel preparation instructions).¹⁷

SODIUM PHOSPHATE SOLUTIONS

Sodium phosphate is an osmotic laxative that draws water into the bowel lumen to promote colonic cleansing. Retention of water in the lumen of the colon stimulates peristalsis and bowel movements.

Advantages and disadvantages of sodium phosphate solutions

Sodium phosphate is widely used worldwide and has been found to be a very acceptable and effective bowel cleansing agent. A recent systematic review of 25 studies¹⁸ found that sodium phosphate was superior to polyethylene glycol in 14 studies, that there was no significant difference in 10 studies, and that only one study found polyethylene glycol to be better tolerated than sodium phosphate.¹⁸ Similarly, a meta-analysis¹⁹ found sodium phosphate to be more effective than polyethylene glycol in bowel cleansing (odds ratio 0.75; P = .0004); more easily completed by patients (odds ratio 0.16; P < .00001); and comparable in terms of adverse events (odds ratio 0.98; P = .81).¹⁹ However, most of the clinical trials excluded patients who had renal failure, ascites, or serious heart disease—the groups most at risk of significant adverse effects from sodium phosphate use. The main reasons sodium phosphate was better tolerated were better flavor and smaller volume (1.5–2 L compared with 4 L for polyethylene glycol).^20–22

The main disadvantage of sodium phosphate is its potential to cause large fluid and electrolyte shifts. Its use has been associated with a variety of electrolyte abnormalities, including hyperphosphatemia, hypocalcemia, hypokalemia, increased plasma osmolality, hyponatremia, and, conversely, hypernatremia.^7,8,23 Asymptomatic hyperphosphatemia alone can be seen in as many as 40% of healthy patients completing sodium phosphate preparations. It may be significant in patients with renal failure and can lead to acute phosphate nephropathy.

Rare adverse events such as nephrocalcinosis with acute renal failure also have been reported, especially in patients taking angiotensin-converting enzyme (ACE) inhibitors or angiotensin receptor blockers.²³

The significant volume contraction and consequent dehydration seen in some patients using sodium phosphate may be decreased by encouraging patients to drink fluids liberally, especially before the day of the procedure and after the procedure.⁷

Recently, renal failure due to hyperphosphatemia (acute phosphate nephropathy) has been reported even in patients with normal kidney function.²⁴ Because of the risk of inappropriate use or overdose associated with over-the counter sodium phosphate, the FDA recommended on December 11, 2008, that sodium phosphate products be available only by prescription when they are used for bowel cleansing.²⁵ The C.B. Fleet Company voluntarily recalled its oral sodium phosphate products sold over the counter (Fleet Phospho-Soda and Fleet EZ-PREP). In addition, the FDA required a black box warning on the prescription oral sodium phosphate products Visicol and OsmoPrep, alerting consumers to the risk of acute phosphate nephropathy.²⁵ According to the FDA, health professionals should use caution when prescribing Visicol or OsmoPrep for patients who may be at higher risk of kidney injury, such as:

• Patients over 55 years of age
• Patients who are dehydrated or who have kidney disease, acute colitis, or delayed bowel emptying
• Patients taking certain drugs that affect kidney function, such as diuretics, ACE inhibitors, angiotensin receptor blockers, and nonsteroidal anti-inflammatory drugs.¹⁶

Commercially available sodium phosphate products

Sodium phosphate products can still be prescribed, but they are no longer available over the counter in the United States. Patients should be screened to make sure they can safely take these products, and the doses should not exceed the maximum recommended.

Figure 2 shows a simplified algorithm for selecting the optimal bowel preparation agent for an individual patient.

OTHER BOWEL PREPARATION AGENTS AND ADJUNCTS

Magnesium citrate

Like sodium phosphate, magnesium citrate is a hyperosmotic agent that promotes bowel cleansing by increasing intraluminal fluid volume. Since magnesium is eliminated solely by the kidney, it should be used with extreme caution in patients with renal insufficiency or renal failure.

Adding magnesium citrate as an adjunct to polyethylene glycol has been shown to reduce the amount of polyethylene glycol solution required (2 L) for the same result.¹⁷

For patients who cannot tolerate polyethylene glycol, a reasonable alternative is magnesium citrate (1 bottle, around 300 mL) the evening before the procedure plus either bisacodyl tablets at the same time as the magnesium citrate or rectal pulsed irrigation immediately before the procedure.⁷

Saline laxatives that include sodium picosulfate and magnesium citrate in combination are available primarily in the United Kingdom for bowel preparation for colonoscopy. Sodium picosulfate acts locally in the colon as a stimulant laxative and by increasing the force of laxatives, whereas magnesium citrate acts as an osmotic laxative by retaining fluids in the colon to clear the colon and rectum of fecal contents. The combination has been found to have similar efficacy and tolerability as sodium phosphate but is not currently available in the United States.²⁶

Enemas

Enemas are sufficient for flexible sigmoidoscopy, but when used alone they do not clean out the proximal colon enough for adequate visualization during colonoscopy. They are best used as adjuncts to other bowel preparation agents when patients present with poor distal colon preparation for colonoscopy.^7,27 Enemas are also useful in washing out the distal segment of bowel in patients with a proximal stoma. The common types of enemas used are tap water, sodium biphosphate (Fleet), and mineral oil.

Tap water enemas distend the rectum and mimic the natural distention by the stool to allow the rectum to empty itself. Tap water (1 L) has fewer adverse effects than sodium biphosphate or mineral oil but is less effective.

Sodium biphosphate (Fleet) enemas draw fluid into the bowel by osmotic action, prompting contraction. One or two bottles are commonly used for bowel cleansing before sigmoidoscopy. However, as with oral sodium phosphate, sodium biphosphate enemas should be avoided in the elderly and in those with renal failure because of the risk of hyperphosphatemia and subsequent hypocalcemia.

In a head-to-head comparison,²⁸ sodium biphosphate enema was found to provide significantly better bowel preparation than the sodium picosulfate-magnesium citrate combination (currently not available in the United States) for flexible sigmoidoscopy, being judged adequate or better in 93% of procedures as opposed to 74%.²⁸

Oil-based enemas such as cottonseed oil plus docusate (Colace) and diatrizoate sodium (Hypaque) are powerful lubricant laxatives that work by slowing the absorption of water from the bowel, so that the stool is softer. However, they have a number of adverse effects, such as severe allergic reactions (including angioedema and anaphylaxis), muscle cramps, and sporadic seepage that can soil the patient's undergarments for up to 24 hours. Also, their safety in children less than 2 years of age and in pregnant and breastfeeding mothers is not established.

Oil-based enemas are usually reserved for short-term use in refractory constipation, especially to soften feces that has become hardened within the rectum (as in fecal impaction).²⁷

Adjuncts

Diet. Dietary modifications alone, such as a clear liquid diet, are inadequate for colonoscopy, but they may be beneficial as adjuncts to other cleansing methods by decreasing the formation of solid residue. Clear liquids also help maintain adequate hydration during bowel preparation and are recommended with all bowel preparation regimens.

Hyperosmolar or stimulant laxatives. Bisacodyl (two to four tablets of 5 mg each), magnesium citrate (one bottle, about 300 mL), and low-dose senna (36 mg, about four 8.6-mg Sennakot tablets) have been used as adjuncts to low-volume polyethylene glycol solution, achieving results similar to those with full-volume polyethylene glycol. Depending on the type of study to be done, these agents are taken within 2 to 6 hours of starting the polyethylene glycol solution.

In contrast, the routine use of nonabsorbable carbohydrates such as mannitol and lactulose is not favored for bowel preparation, since the hydrogen gas produced by bacterial fermentation of the nonabsorbed carbohydrates increases the risk of explosion during electrosurgical procedures.²⁹

Antiemetic agents. Metoclopramide (5–10 mg), a dopamine antagonist gastroprokinetic that sensitizes tissues to the action of acetylcholine, is commonly used to prevent nausea or vomiting associated with bowel preparation agents.^7,30

Antifoaming agent. Simethicone (three tablets of 80 mg each, total dose 240 mg), an anti-flatulent, anti-gas agent, is prescribed by many gastroenterologists in an attempt to reduce bubbles during colonoscopy and improve visibility. It works by reducing the surface tension of air bubbles and causing small bubbles to coalesce into larger ones that pass more easily with belching or flatulence.

Nasogastric or orogastric tubes have been used to instill colonic preparations, especially for inpatients unable to drink polyethylene glycol solutions or for patients who are unresponsive or mechanically ventilated. This method can also be useful for rapid bowel cleansing (within 2 to 3 hours) for patients with lower gastrointestinal bleeding. However, routine use of a nasogastric tube solely for bowel preparation is discouraged as it can lead to severe complications, such as aspiration and trauma during insertion.⁷

OTHER CONSIDERATIONS

Patient education

The importance of patient education for successful bowel preparation cannot be overemphasized. Patients need to be informed about why they need to undergo colonoscopy, the importance of bowel preparation, the side effects of agents used, and the exact preparation instructions. An interactive educational tutorial about colonoscopy for patients is available at Medline Plus at http://www.nlm.nih.gov/medlineplus/tutorials/colonoscopy/htm/index.htm.

Role of hydration

A commonly held misconception is that patients taking 4 L of polyethylene glycol do not need additional hydration, since they are already ingesting such a large volume of fluid. Given that bowel preparations induce diarrhea and, in some instances, nausea and vomiting, all patients taking bowel preparations are at risk of dehydration.³² In fact, the fluid loss during bowel preparation may exceed 2 to 3 L. It is not surprising that many safety issues associated with bowel preparation agents are related to dehydration and its complications.

Hence, patients should be advised to consume at least 64 oz (approximately 2 L) of clear fluid on the day before the colonoscopy. According to the American Society of Anesthesiologists, clear liquids can be safely ingested up until 2 hours before receiving anesthesia.³³ Patients should contact their physicians if they experience vomiting or cannot comply with clear liquid volume instructions prior to colonoscopy. Metoclopramide has been found useful in many cases of nausea or vomiting associated with bowel preparation agents.¹⁸ In addition, patients should also be reminded to keep drinking extra fluids after the procedure is completed to reduce the risk of dehydration and its complications (Table 2).