Applied Evidence

Lucy’s parents were having difficulty reconciling the child they knew at home with the one they sent to school. At home, the 7-year-old spoke easily with her parents and siblings. At school and in other public places such as the supermarket, mall, or park, she would not speak at all. This had been going on for several years.

While Lucy was attending school without difficulty, she seemed sullen and withdrawn to her teachers. Lucy had friends at school, but would communicate with them by writing messages on paper or drawing letters in the air. Although Lucy had passed first grade on the basis of her written work, her parents worried about her ability to succeed in second grade, where her teachers expected her to participate in class and she would need to take standardized tests that required audible verbal responses. At the urging of a school counselor, Lucy’s parents took her to their family physician. They needed help in drawing her out so that the rest of the world could get to know the Lucy that they knew and loved.

Lucy’s case is typical of selective mutism

Lucy was suffering from selective mutism, the persistent failure to speak in specific social situations where speaking is expected, such as at school and with playmates.¹ Lucy’s case is typical in that children with selective mutism speak well in other situations, typically at home. Thus, the disorder is not due to a communication disorder such as stuttering and it is not due to a lack of knowledge or comfort with language.

To meet diagnostic criteria, the disorder must last at least 1 month, though it can last for several years, and must interfere with a child’s education or ability to communicate socially.¹

A little known disorder makes national headlines

Selective mutism occurs in 0.2%–2.0% of children, affects boys and girls equally, and often begins at 3 to 6 years of age.^2-9 The disorder gained national attention this past spring when it was revealed that the shooter in the Virginia Tech massacre—Seung-Hui Cho—had, as an adolescent, been diagnosed with selective mutism.¹⁰ Though his diagnosis made plenty of headlines, the disorder itself occupies little space in the pediatric literature.¹¹

What we do know is that selective mutism can have a chronic course that affects a child’s ability to form friendships, complete academic tasks, develop appropriate language and social skills, and participate in standardized testing.^2-9 While parents often attribute their child’s behavior to shyness, this disorder goes beyond that. While shy children function, those with selective mutism struggle socially, emotionally, and academically. Children with selective mutism say the words won’t come out and their body won’t let them speak. One father of a 7-year-old girl with selective mutism said that his daughter “describes it as the words get stuck in her toes.”¹²

Many researchers theorize that children with selective mutism have severe social anxiety, and assessment and treatment strategies are typically based on this notion. Reports in the literature have also suggested that selective mutism is related to a developmental disorder or delay, as seen in autistic spectrum disorders, as well as anxiety disorders and depression.¹³ In addition, selective mutism has been linked to oppositional defiant disorder and subtle language impairments.^14,15

Begin by excluding competing explanations

If a family is referred to you for possible selective mutism, you’ll first need to exclude competing explanations for the problem, such as hearing difficulties, speech and language disorders, school-based threats, and medical problems, such as asthma, which could prevent a child from speaking comfortably in a public setting. Assuming you are able to exclude these, and other competing explanations, you’ll need to ask 8 questions of parents and school officials. The answers you get will help you to screen for selective mutism and identify key antecedents and consequences to the behavior that are important in addressing the problem.

8 questions you’ll need to ask

1. What specific settings involve failure to speak? Children with selective mutism typically have difficulties in school, on the playground, in malls, and at restaurants.

2. Has the mutism lasted at least one month? Brief refusal to speak is not uncommon in children.

3. Does the child speak well at home with people she knows well? Most children with selective mutism speak well at home with family members, which belies a communication disorder.

4. Is failure to speak significantly interfering with the child’s academic or social development? Selective mutism must involve a significant interference in daily functioning.

5. What circumstances surround each episode of mutism? In particular, is the mutism associated with the desire to increase social attention or stimulation from others, decrease anxiety, or avoid or withdraw from adult commands or requests? A full understanding of selective mutism must include a review of triggers like these.

6. Can the child be encouraged to speak audibly in any way in certain public settings? Children with selective mutism who can speak to some extent in public situations may have a better prognosis than those who do not.⁹

7. How do others respond to, or compensate for, the child’s mutism? Do they complete tasks for the child? Order food for her? Allow whispers or communication via writing? Excessive accommodations enable a child with selective mutism to maintain the behavior.

8. Does the child appear anxious or depressed in situations involving mutism? Mutism is often linked to poor affect and social anxiety.

Still suspect selective mutism? Make a referral

If after asking these 8 questions you still suspect selective mutism, you will need to make a referral to a child psychologist or other mental health professional who specializes in behavioral strategies to treat selective mutism. The psychologist will meet with the child and utilize more formal methods of assessment.

The psychologist may use the Anxiety Disorders Interview Schedule for DSM-IV-TR (child and parent versions), a structured interview that emphasizes anxiety-based disorders and includes a section for selective mutism based on DSM-IV-TR criteria.¹⁶ Another helpful tool is the Functional Diagnostic Profile for Selective Mutism, which helps to assess contextual factors that surround refusal to speak.¹⁷ The challenge, of course, in conducting these interviews is that many children will not answer the psychologist’s questions verbally, but may communicate via nonverbal gestures, writing, or other creative ways such as drawing letters in the air.

Parent and teacher logs provide valuable insight

Ongoing behavioral observations or daily logbooks completed by parents and teachers are also important in treating a child with selective mutism. If parents aren’t doing so already at the suggestion of the therapist, ask them to keep a logbook and encourage them to ask the child’s teachers to keep one, as well. This log should include the number and volume of words spoken, to whom the child has spoken, where the child spoke, how others reacted (ie, Did they pay extra attention to the child?), and the child’s compensatory behaviors such as whispering, pointing, crying, nodding, or frowning.¹⁸ Parents and teachers should also note the antecedents to mutism, including demands (or requests) from others, or the presence of unfamiliar people.

A focus on systematic desensitization

Much of the literature regarding the treatment of selective mutism consists of case reports and small-scale studies.¹⁹ The paucity of data is, in part, a reflection of the rarity of the problem as well as a historical lack of consensus among researchers about the nature and diagnostic criteria for selective mutism.

The studies we do have on selective mutism indicate that behavior modification can be effective for these patients. Behavioral treatment primarily involves child- and parent-based practices.²⁰

Child-based practices include management of physical sensations of anxiety often via relaxation training and breathing retraining. Neither technique requires verbal input from a child. In addition, the therapist will have the child practice audible words in progressively more difficult settings via systematic desensitization. These settings often include the following, in order: the child’s home (with the therapist), the therapist’s office, community settings such as a mall or a restaurant, and finally, school.

In these settings, the therapist might, for example, encourage the patient to:

• speak on the telephone or answer the door at home.
• stay at the therapist’s office until at least one word is audibly spoken. (One approach that the therapist may use to increase rapport and provide opportunities to prompt speech is to play games with the child.)
• order her own food at a local restaurant.
• read a story aloud to classmates at school.

In addition to systematic desensitization, the therapist may use audio or videotapes to help the youngster. In one scenario, the therapist may ask the parents to bring in an audiotape or videotape of the child at home when she is speaking. The therapist then provides ample reinforcement for the child’s speech by praising the child for having a beautiful voice and for making the tape. Audio and videotapes may also be used to help the child speak louder or work on her articulation.

Reinforce the positive, avoid “shaming”

Parent-based practices include reinforcing the child’s speech in different settings, developing new and independent exposures to generalize speech, and consulting with school officials on an ongoing basis to address social and academic problems from poor use of speech.

The therapist is likely to remind the parent that positive reinforcement is key, while punishing or “shaming” the child is ineffective.¹¹ Some therapists will make use of a technique called “token economy,” where the child’s progress earns her colored chips that can be redeemed for a trip to a video game center, or a special treat, such as a sleep-over with a trusted friend.¹¹

With the therapist’s help, parents can also model appropriate social interactions for a child with selective mutism because social skills may not be sufficiently developed. For example, the parent might encourage the child to answer the door or telephone. The parent might also monitor the child’s interactions with friends and provide feedback. In addition, the therapist may encourage the parents to ignore inappropriate compensatory behaviors as treatment progresses. Instead, a child with selective mutism may be increasingly challenged to provide audible speech to make requests or otherwise communicate.²¹

Drug therapy may include MAOIs and SSRIs

Medical treatment for selective mutism may include monoamine oxidase inhibitors (MAOIs) such as phenelzine (30–60 mg/day) or selective serotonin reuptake inhibitors (SSRIs) such as fluoxetine (10–60 mg/day), sertraline (100 mg/day), fluvoxamine (50–100 mg/day), citalopram (20–40 mg/day), or paroxetine (5 mg/day).²² However, no large-scale studies of pharmacotherapy for selective mutism have been conducted and medication effects are quite variable. The literature in this area has consisted largely of case studies.

After 9 months of treatment, Lucy speaks up at school

Lucy’s therapist (JLV) utilized many of the behavioral procedures described in this article. Treatment included exposure-based practices at home, in a clinic, in various community settings, and at school. By the end of her 9 months of treatment, Lucy was speaking independently to her classmates and teachers, though ongoing praise by her teachers was needed to encourage her to maintain an appropriate speech volume.

The literature on selective mutism suggests Lucy’s case was not unusual; it typically takes several months for these young patients to improve. Ongoing communication issues, though, often linger.⁹

Correspondence
Christopher A. Kearney, PhD, Department of Psychology, University of Nevada, Las Vegas, 4505 maryland Parkway, Las Vegas, Nv 89154-5030; [email protected]

Lucy’s parents were having difficulty reconciling the child they knew at home with the one they sent to school. At home, the 7-year-old spoke easily with her parents and siblings. At school and in other public places such as the supermarket, mall, or park, she would not speak at all. This had been going on for several years.

While Lucy was attending school without difficulty, she seemed sullen and withdrawn to her teachers. Lucy had friends at school, but would communicate with them by writing messages on paper or drawing letters in the air. Although Lucy had passed first grade on the basis of her written work, her parents worried about her ability to succeed in second grade, where her teachers expected her to participate in class and she would need to take standardized tests that required audible verbal responses. At the urging of a school counselor, Lucy’s parents took her to their family physician. They needed help in drawing her out so that the rest of the world could get to know the Lucy that they knew and loved.

Lucy’s case is typical of selective mutism

Lucy was suffering from selective mutism, the persistent failure to speak in specific social situations where speaking is expected, such as at school and with playmates.¹ Lucy’s case is typical in that children with selective mutism speak well in other situations, typically at home. Thus, the disorder is not due to a communication disorder such as stuttering and it is not due to a lack of knowledge or comfort with language.

To meet diagnostic criteria, the disorder must last at least 1 month, though it can last for several years, and must interfere with a child’s education or ability to communicate socially.¹

A little known disorder makes national headlines

Selective mutism occurs in 0.2%–2.0% of children, affects boys and girls equally, and often begins at 3 to 6 years of age.^2-9 The disorder gained national attention this past spring when it was revealed that the shooter in the Virginia Tech massacre—Seung-Hui Cho—had, as an adolescent, been diagnosed with selective mutism.¹⁰ Though his diagnosis made plenty of headlines, the disorder itself occupies little space in the pediatric literature.¹¹

What we do know is that selective mutism can have a chronic course that affects a child’s ability to form friendships, complete academic tasks, develop appropriate language and social skills, and participate in standardized testing.^2-9 While parents often attribute their child’s behavior to shyness, this disorder goes beyond that. While shy children function, those with selective mutism struggle socially, emotionally, and academically. Children with selective mutism say the words won’t come out and their body won’t let them speak. One father of a 7-year-old girl with selective mutism said that his daughter “describes it as the words get stuck in her toes.”¹²

Many researchers theorize that children with selective mutism have severe social anxiety, and assessment and treatment strategies are typically based on this notion. Reports in the literature have also suggested that selective mutism is related to a developmental disorder or delay, as seen in autistic spectrum disorders, as well as anxiety disorders and depression.¹³ In addition, selective mutism has been linked to oppositional defiant disorder and subtle language impairments.^14,15

Begin by excluding competing explanations

If a family is referred to you for possible selective mutism, you’ll first need to exclude competing explanations for the problem, such as hearing difficulties, speech and language disorders, school-based threats, and medical problems, such as asthma, which could prevent a child from speaking comfortably in a public setting. Assuming you are able to exclude these, and other competing explanations, you’ll need to ask 8 questions of parents and school officials. The answers you get will help you to screen for selective mutism and identify key antecedents and consequences to the behavior that are important in addressing the problem.

8 questions you’ll need to ask

1. What specific settings involve failure to speak? Children with selective mutism typically have difficulties in school, on the playground, in malls, and at restaurants.

2. Has the mutism lasted at least one month? Brief refusal to speak is not uncommon in children.

3. Does the child speak well at home with people she knows well? Most children with selective mutism speak well at home with family members, which belies a communication disorder.

4. Is failure to speak significantly interfering with the child’s academic or social development? Selective mutism must involve a significant interference in daily functioning.

5. What circumstances surround each episode of mutism? In particular, is the mutism associated with the desire to increase social attention or stimulation from others, decrease anxiety, or avoid or withdraw from adult commands or requests? A full understanding of selective mutism must include a review of triggers like these.

6. Can the child be encouraged to speak audibly in any way in certain public settings? Children with selective mutism who can speak to some extent in public situations may have a better prognosis than those who do not.⁹

7. How do others respond to, or compensate for, the child’s mutism? Do they complete tasks for the child? Order food for her? Allow whispers or communication via writing? Excessive accommodations enable a child with selective mutism to maintain the behavior.

8. Does the child appear anxious or depressed in situations involving mutism? Mutism is often linked to poor affect and social anxiety.

Still suspect selective mutism? Make a referral

If after asking these 8 questions you still suspect selective mutism, you will need to make a referral to a child psychologist or other mental health professional who specializes in behavioral strategies to treat selective mutism. The psychologist will meet with the child and utilize more formal methods of assessment.

The psychologist may use the Anxiety Disorders Interview Schedule for DSM-IV-TR (child and parent versions), a structured interview that emphasizes anxiety-based disorders and includes a section for selective mutism based on DSM-IV-TR criteria.¹⁶ Another helpful tool is the Functional Diagnostic Profile for Selective Mutism, which helps to assess contextual factors that surround refusal to speak.¹⁷ The challenge, of course, in conducting these interviews is that many children will not answer the psychologist’s questions verbally, but may communicate via nonverbal gestures, writing, or other creative ways such as drawing letters in the air.

Parent and teacher logs provide valuable insight

Ongoing behavioral observations or daily logbooks completed by parents and teachers are also important in treating a child with selective mutism. If parents aren’t doing so already at the suggestion of the therapist, ask them to keep a logbook and encourage them to ask the child’s teachers to keep one, as well. This log should include the number and volume of words spoken, to whom the child has spoken, where the child spoke, how others reacted (ie, Did they pay extra attention to the child?), and the child’s compensatory behaviors such as whispering, pointing, crying, nodding, or frowning.¹⁸ Parents and teachers should also note the antecedents to mutism, including demands (or requests) from others, or the presence of unfamiliar people.

A focus on systematic desensitization

Much of the literature regarding the treatment of selective mutism consists of case reports and small-scale studies.¹⁹ The paucity of data is, in part, a reflection of the rarity of the problem as well as a historical lack of consensus among researchers about the nature and diagnostic criteria for selective mutism.

The studies we do have on selective mutism indicate that behavior modification can be effective for these patients. Behavioral treatment primarily involves child- and parent-based practices.²⁰

Child-based practices include management of physical sensations of anxiety often via relaxation training and breathing retraining. Neither technique requires verbal input from a child. In addition, the therapist will have the child practice audible words in progressively more difficult settings via systematic desensitization. These settings often include the following, in order: the child’s home (with the therapist), the therapist’s office, community settings such as a mall or a restaurant, and finally, school.

In these settings, the therapist might, for example, encourage the patient to:

• speak on the telephone or answer the door at home.
• stay at the therapist’s office until at least one word is audibly spoken. (One approach that the therapist may use to increase rapport and provide opportunities to prompt speech is to play games with the child.)
• order her own food at a local restaurant.
• read a story aloud to classmates at school.

In addition to systematic desensitization, the therapist may use audio or videotapes to help the youngster. In one scenario, the therapist may ask the parents to bring in an audiotape or videotape of the child at home when she is speaking. The therapist then provides ample reinforcement for the child’s speech by praising the child for having a beautiful voice and for making the tape. Audio and videotapes may also be used to help the child speak louder or work on her articulation.

Reinforce the positive, avoid “shaming”

Parent-based practices include reinforcing the child’s speech in different settings, developing new and independent exposures to generalize speech, and consulting with school officials on an ongoing basis to address social and academic problems from poor use of speech.

The therapist is likely to remind the parent that positive reinforcement is key, while punishing or “shaming” the child is ineffective.¹¹ Some therapists will make use of a technique called “token economy,” where the child’s progress earns her colored chips that can be redeemed for a trip to a video game center, or a special treat, such as a sleep-over with a trusted friend.¹¹

With the therapist’s help, parents can also model appropriate social interactions for a child with selective mutism because social skills may not be sufficiently developed. For example, the parent might encourage the child to answer the door or telephone. The parent might also monitor the child’s interactions with friends and provide feedback. In addition, the therapist may encourage the parents to ignore inappropriate compensatory behaviors as treatment progresses. Instead, a child with selective mutism may be increasingly challenged to provide audible speech to make requests or otherwise communicate.²¹

Drug therapy may include MAOIs and SSRIs

Medical treatment for selective mutism may include monoamine oxidase inhibitors (MAOIs) such as phenelzine (30–60 mg/day) or selective serotonin reuptake inhibitors (SSRIs) such as fluoxetine (10–60 mg/day), sertraline (100 mg/day), fluvoxamine (50–100 mg/day), citalopram (20–40 mg/day), or paroxetine (5 mg/day).²² However, no large-scale studies of pharmacotherapy for selective mutism have been conducted and medication effects are quite variable. The literature in this area has consisted largely of case studies.

After 9 months of treatment, Lucy speaks up at school

Lucy’s therapist (JLV) utilized many of the behavioral procedures described in this article. Treatment included exposure-based practices at home, in a clinic, in various community settings, and at school. By the end of her 9 months of treatment, Lucy was speaking independently to her classmates and teachers, though ongoing praise by her teachers was needed to encourage her to maintain an appropriate speech volume.

The literature on selective mutism suggests Lucy’s case was not unusual; it typically takes several months for these young patients to improve. Ongoing communication issues, though, often linger.⁹

Correspondence
Christopher A. Kearney, PhD, Department of Psychology, University of Nevada, Las Vegas, 4505 maryland Parkway, Las Vegas, Nv 89154-5030; [email protected]

Lucy’s parents were having difficulty reconciling the child they knew at home with the one they sent to school. At home, the 7-year-old spoke easily with her parents and siblings. At school and in other public places such as the supermarket, mall, or park, she would not speak at all. This had been going on for several years.

While Lucy was attending school without difficulty, she seemed sullen and withdrawn to her teachers. Lucy had friends at school, but would communicate with them by writing messages on paper or drawing letters in the air. Although Lucy had passed first grade on the basis of her written work, her parents worried about her ability to succeed in second grade, where her teachers expected her to participate in class and she would need to take standardized tests that required audible verbal responses. At the urging of a school counselor, Lucy’s parents took her to their family physician. They needed help in drawing her out so that the rest of the world could get to know the Lucy that they knew and loved.

Lucy’s case is typical of selective mutism

Lucy was suffering from selective mutism, the persistent failure to speak in specific social situations where speaking is expected, such as at school and with playmates.¹ Lucy’s case is typical in that children with selective mutism speak well in other situations, typically at home. Thus, the disorder is not due to a communication disorder such as stuttering and it is not due to a lack of knowledge or comfort with language.

To meet diagnostic criteria, the disorder must last at least 1 month, though it can last for several years, and must interfere with a child’s education or ability to communicate socially.¹

A little known disorder makes national headlines

Selective mutism occurs in 0.2%–2.0% of children, affects boys and girls equally, and often begins at 3 to 6 years of age.^2-9 The disorder gained national attention this past spring when it was revealed that the shooter in the Virginia Tech massacre—Seung-Hui Cho—had, as an adolescent, been diagnosed with selective mutism.¹⁰ Though his diagnosis made plenty of headlines, the disorder itself occupies little space in the pediatric literature.¹¹

What we do know is that selective mutism can have a chronic course that affects a child’s ability to form friendships, complete academic tasks, develop appropriate language and social skills, and participate in standardized testing.^2-9 While parents often attribute their child’s behavior to shyness, this disorder goes beyond that. While shy children function, those with selective mutism struggle socially, emotionally, and academically. Children with selective mutism say the words won’t come out and their body won’t let them speak. One father of a 7-year-old girl with selective mutism said that his daughter “describes it as the words get stuck in her toes.”¹²

Many researchers theorize that children with selective mutism have severe social anxiety, and assessment and treatment strategies are typically based on this notion. Reports in the literature have also suggested that selective mutism is related to a developmental disorder or delay, as seen in autistic spectrum disorders, as well as anxiety disorders and depression.¹³ In addition, selective mutism has been linked to oppositional defiant disorder and subtle language impairments.^14,15

Begin by excluding competing explanations

If a family is referred to you for possible selective mutism, you’ll first need to exclude competing explanations for the problem, such as hearing difficulties, speech and language disorders, school-based threats, and medical problems, such as asthma, which could prevent a child from speaking comfortably in a public setting. Assuming you are able to exclude these, and other competing explanations, you’ll need to ask 8 questions of parents and school officials. The answers you get will help you to screen for selective mutism and identify key antecedents and consequences to the behavior that are important in addressing the problem.

8 questions you’ll need to ask

1. What specific settings involve failure to speak? Children with selective mutism typically have difficulties in school, on the playground, in malls, and at restaurants.

2. Has the mutism lasted at least one month? Brief refusal to speak is not uncommon in children.

3. Does the child speak well at home with people she knows well? Most children with selective mutism speak well at home with family members, which belies a communication disorder.

4. Is failure to speak significantly interfering with the child’s academic or social development? Selective mutism must involve a significant interference in daily functioning.

5. What circumstances surround each episode of mutism? In particular, is the mutism associated with the desire to increase social attention or stimulation from others, decrease anxiety, or avoid or withdraw from adult commands or requests? A full understanding of selective mutism must include a review of triggers like these.

6. Can the child be encouraged to speak audibly in any way in certain public settings? Children with selective mutism who can speak to some extent in public situations may have a better prognosis than those who do not.⁹

7. How do others respond to, or compensate for, the child’s mutism? Do they complete tasks for the child? Order food for her? Allow whispers or communication via writing? Excessive accommodations enable a child with selective mutism to maintain the behavior.

8. Does the child appear anxious or depressed in situations involving mutism? Mutism is often linked to poor affect and social anxiety.

Still suspect selective mutism? Make a referral

If after asking these 8 questions you still suspect selective mutism, you will need to make a referral to a child psychologist or other mental health professional who specializes in behavioral strategies to treat selective mutism. The psychologist will meet with the child and utilize more formal methods of assessment.

The psychologist may use the Anxiety Disorders Interview Schedule for DSM-IV-TR (child and parent versions), a structured interview that emphasizes anxiety-based disorders and includes a section for selective mutism based on DSM-IV-TR criteria.¹⁶ Another helpful tool is the Functional Diagnostic Profile for Selective Mutism, which helps to assess contextual factors that surround refusal to speak.¹⁷ The challenge, of course, in conducting these interviews is that many children will not answer the psychologist’s questions verbally, but may communicate via nonverbal gestures, writing, or other creative ways such as drawing letters in the air.

Parent and teacher logs provide valuable insight

Ongoing behavioral observations or daily logbooks completed by parents and teachers are also important in treating a child with selective mutism. If parents aren’t doing so already at the suggestion of the therapist, ask them to keep a logbook and encourage them to ask the child’s teachers to keep one, as well. This log should include the number and volume of words spoken, to whom the child has spoken, where the child spoke, how others reacted (ie, Did they pay extra attention to the child?), and the child’s compensatory behaviors such as whispering, pointing, crying, nodding, or frowning.¹⁸ Parents and teachers should also note the antecedents to mutism, including demands (or requests) from others, or the presence of unfamiliar people.

A focus on systematic desensitization

Much of the literature regarding the treatment of selective mutism consists of case reports and small-scale studies.¹⁹ The paucity of data is, in part, a reflection of the rarity of the problem as well as a historical lack of consensus among researchers about the nature and diagnostic criteria for selective mutism.

The studies we do have on selective mutism indicate that behavior modification can be effective for these patients. Behavioral treatment primarily involves child- and parent-based practices.²⁰

Child-based practices include management of physical sensations of anxiety often via relaxation training and breathing retraining. Neither technique requires verbal input from a child. In addition, the therapist will have the child practice audible words in progressively more difficult settings via systematic desensitization. These settings often include the following, in order: the child’s home (with the therapist), the therapist’s office, community settings such as a mall or a restaurant, and finally, school.

In these settings, the therapist might, for example, encourage the patient to:

• speak on the telephone or answer the door at home.
• stay at the therapist’s office until at least one word is audibly spoken. (One approach that the therapist may use to increase rapport and provide opportunities to prompt speech is to play games with the child.)
• order her own food at a local restaurant.
• read a story aloud to classmates at school.

In addition to systematic desensitization, the therapist may use audio or videotapes to help the youngster. In one scenario, the therapist may ask the parents to bring in an audiotape or videotape of the child at home when she is speaking. The therapist then provides ample reinforcement for the child’s speech by praising the child for having a beautiful voice and for making the tape. Audio and videotapes may also be used to help the child speak louder or work on her articulation.

Reinforce the positive, avoid “shaming”

Parent-based practices include reinforcing the child’s speech in different settings, developing new and independent exposures to generalize speech, and consulting with school officials on an ongoing basis to address social and academic problems from poor use of speech.

The therapist is likely to remind the parent that positive reinforcement is key, while punishing or “shaming” the child is ineffective.¹¹ Some therapists will make use of a technique called “token economy,” where the child’s progress earns her colored chips that can be redeemed for a trip to a video game center, or a special treat, such as a sleep-over with a trusted friend.¹¹

With the therapist’s help, parents can also model appropriate social interactions for a child with selective mutism because social skills may not be sufficiently developed. For example, the parent might encourage the child to answer the door or telephone. The parent might also monitor the child’s interactions with friends and provide feedback. In addition, the therapist may encourage the parents to ignore inappropriate compensatory behaviors as treatment progresses. Instead, a child with selective mutism may be increasingly challenged to provide audible speech to make requests or otherwise communicate.²¹

Drug therapy may include MAOIs and SSRIs

Medical treatment for selective mutism may include monoamine oxidase inhibitors (MAOIs) such as phenelzine (30–60 mg/day) or selective serotonin reuptake inhibitors (SSRIs) such as fluoxetine (10–60 mg/day), sertraline (100 mg/day), fluvoxamine (50–100 mg/day), citalopram (20–40 mg/day), or paroxetine (5 mg/day).²² However, no large-scale studies of pharmacotherapy for selective mutism have been conducted and medication effects are quite variable. The literature in this area has consisted largely of case studies.

After 9 months of treatment, Lucy speaks up at school

Lucy’s therapist (JLV) utilized many of the behavioral procedures described in this article. Treatment included exposure-based practices at home, in a clinic, in various community settings, and at school. By the end of her 9 months of treatment, Lucy was speaking independently to her classmates and teachers, though ongoing praise by her teachers was needed to encourage her to maintain an appropriate speech volume.

The literature on selective mutism suggests Lucy’s case was not unusual; it typically takes several months for these young patients to improve. Ongoing communication issues, though, often linger.⁹

Correspondence
Christopher A. Kearney, PhD, Department of Psychology, University of Nevada, Las Vegas, 4505 maryland Parkway, Las Vegas, Nv 89154-5030; [email protected]

Ann G. is a 34-year-old mother of 2 who had been coming to the office for her annual Pap smear for several years. Her medical history is significant only for her vaginal deliveries and mild GERD. Her medications include oral contraceptive pills (OCPs) and over-the-counter Zantac as needed. on her most recent annual visit, my medical assistant, Tammy, took Ann’s vital signs. The chart had a section about smoking status, and Tammy noted that Ann smoked.

During the office visit, I explained to Ann that her smoking was a serious health risk, and that she needed to quit. She would also need to find a new form of birth control next year, as smoking increases the risks of using OCPs. She nervously laughed off the warning.

The following year, Anne confessed to Tammy that she was still a smoker. Tammy asked her again about quitting. Ann was still adamant: “No way—I can’t do it.” Nonetheless, during the office visit, I brought up the subject of her smoking. She admitted that she was afraid that quitting smoking would cause her to gain weight. I attempted to address her fears, and then talked about other birth control methods to consider. I gave her a 3-month prescription of OCPs, and told her in 3 months we would discuss what she wanted to do about birth control.

Ann faces an uphill battle. The amount of nicotine in cigarettes is increasing,¹ making it harder for her to quit. The good news is that the treatment of tobacco addiction is constantly improving and the number of tools in our arsenal is growing. In fact, there are many resources that we can try before turning to the prescription pad. However, when needed, pharmacotherapy is an important adjunct for achieving abstinence.

“5-A” strategy sets stage for success

The Agency for Healthcare Research and Quality (AHRQ) has published Treating Tobacco Use and Dependence, a useful guide for helping patients quit.^2,3 These guidelines discuss many aspects of tobacco cessation, from counseling to pharmacotherapy to reimbursement issues. The guidelines break down the smoking cessation process into the 5 A’s:

1. Ask each patient about her smoking status.
2. Advise each patient who smokes that she needs to stop smoking.
3. Assess your patient’s willingness to make a quit attempt in the next 30 days.
4. Assist your patient either in making this quit attempt or in motivating her to consider a quit attempt later.
5. Arrange close follow-up of any quit attempts to help prevent relapse.

The Ask and Act program from the American Academy of Family Physicians (AAFP) outlines a similar strategy.⁴ The program instructs physicians to Ask every patient about her tobacco use and to Act to help her quit, via on-or off-site counseling, quitlines, patient education materials, self-help guides or Web sites, cessation classes, and pharmacotherapy.

Take advantage of every opportunity you have to discuss the issue with patients; short conversations can make a difference. A Cochrane Review of 39 trials including 31,000 smokers⁵ revealed that even brief advice—simply encouraging patients to quit—was statistically significant (odds ratio [OR]=1.74; 95% confidence interval [CI], 1.48–2.05). The pooled data generated a quit rate difference of 2.5%: for every 40 people who were told to quit, 1 more smoker would.

Empower the office staff

Enlisting the help of the office staff can have a significant impact on the health of the patients. A proactive approach was studied by Fiore et al.⁶ Medical assistants, while assessing smoking status, invited all smokers to participate in a cessation study. (The assistants received periodic thank-you gifts for their efforts.)

The participants were randomized to either self-selected treatment or nicotine replacement therapy (NRT) patches, with or without a support program. Some who received the patches and support program also received individual counseling. Fiore et al showed that the majority of smokers were open to attempts to quit smoking. The 13% point-prevalence abstinence rate 1 year out is comparable with the rate obtained (14%) with smokers volunteering for NRT studies in the Cochrane review of 39 trials, noted earlier.

Likewise, in a randomized controlled trial (RCT) involving community-based primary care clinics, Katz⁷ demonstrated that intake clinicians could also play an important role in smoking cessation (SOR: A). In the study, researchers trained intake clinicians (including registered nurses, license practical nurses, and medical assistants) to identify smokers, provide brief counseling, and assist in their preparation to quit. Patients were offered vouchers for patches and a counselor’s business card. Intake clinicians received periodic feedback on their performance based on exit interviews of the patients. The researchers found that these interventions had a statistically significant effect in moderate-to-heavy smokers in quit attempts, quit rates, and continuous abstinence.

Our patient has a change of heart

At the 3-month follow up, Tammy learned that Ann was still smoking—but she now wanted to quit. Ann said that she’d found a pack of cigarettes in her 14-year-old daughter’s backpack, and felt that the only way to prevent her from getting hooked was to set a good example.

Tammy gave her the state’s quitline number and suggested some online quitting programs. Tammy worked with Ann to choose her target quit date and to pick the Web-based program she was going to use. Ann said that she liked the idea that she could go online whenever she needed support. She also liked the fact that she could put her quit date into the system, so it would give her timely reminders of all her reasons to quit when she logged on.

I wrote Ann a prescription for varenicline and her OCPs, and told her I wanted to see her in 4 weeks. For her part, Tammy added Ann to her list of patients to call the day after her quit date. Tammy makes this her practice with patients because she knows that one well-timed phone call can be the key to a successful quitting attempt.

Outside support improves abstinence rates

Improving your patients’ chances of long-term abstinence hinges, in part, on making the most of outside support. In many cases, your patients can take advantage of them without leaving their homes.

Outside support improves abstinence rates

Quitlines increase quit rates, decrease relapse Telephone counseling is an effective support system.⁸ Smokers who call to a single number (800-QUITNOW)—a service provided by the National Cancer Institute—are directed to the quitline for their state. Also, smokers can call the National Cancer Institute directly at their quitline (877-44U-QUIT). Calling a quitline provides smokers with real-time counseling and information about how to quit smoking. Quitlines can be appealing to those patients who are uncomfortable discussing their smoking in a group—and it’s free to the patient.

The research supports the use of such help lines. Zhu’s study⁹ of the California Smokers’ Helpline (SOR: A) was a proactive protocol where smokers were funneled into a research trial when the help line was overwhelmed. The smokers in the treatment arm of this RCT were assigned a counselor who called the smokers as many as 6 times, following a relapse-sensitive schedule. The 12-month abstinence rate increased from 4.1% to 7.5% (P<.001) in the group that had close telephone contact. This improved quit rate reflected both an increase of percentage of smokers who quit and, more importantly, a decrease in quitters who relapsed.

Another prospective RCT¹⁰ (SOR: A) enrolled patients from Veterans Affairs (VA) medical centers and involved the same proactive telephone protocol as Zhu used. The treatment group was offered telephone counseling as well as pharmacotherapy; the control group had access to the regular smoking cessation program of the VA system. Regardless of which group an individual was assigned, if that participant used both the counseling and the pharmacotherapy, the quit rate was similar: control (12.7%) and treatment (11.9%). However, only 18% of the controls used both services. The treatment group accessed the combined programs of counseling and medications at a rate of 88%. This led to the difference in 6-month abstinence rates of 13.0% in the treatment group and 4.1% in the control group (OR=3.50; 95% CI, 1.99–6.15). Patients who were directed to and enrolled in treatment programs were therefore more likely to attempt to quit and remain abstinent for up to 6 months.

Web-based programs offer reminders

Like quitlines, Web-based programs offer smokers immediate feedback to help them quit. Many of the programs include links to quitting resources, stories from former smokers and cancer patients, live advice from counselors, and message boards (TABLE 1). Web-based programs have been shown to help improve quit rates.

One study¹¹ compared 2 Web programs involving 11,969 smokers. This RCT (SOR: B) looked at an interactive program based, in part, on the AHRQ treatment guidelines. This program generates personalized letters for the participants along with monthly e-mail reminders. A modified program was used as the control. The control program was developed by a maker of NRT products, and contained more information about nicotine than about tobacco dependence and cessation. This program was also shorter than the interactive program, which was designed to assist smoking cessation.

Both programs improved quit rates: 10.9% for the interactive program and 8% for the modified/control program, compared with 3.3% for no treatment at all. Although this study was based on participant reports of abstinence over the previous 7 days, and had low follow-up rates (which Internet studies tend to have), the interactive program produced 1 more quitter for every 26 participants than the modified (control) program, using an intent-to-treat analysis (14.6% vs. 10.7%, P<.001, OR=1.43, 95% CI, 1.28–1.59).

Another study¹² looked at the use of a more extensive Web site, combining video, audio, and text. This RCT (SOR: B) was fully automated and delivered entirely by computer. Again, using the AHRQ guidelines and other sources, researchers designed a series of 5 modules to simulate working with a live counselor. There were 13 different versions, to match the demographics of the participant. The modules ended with a “quit calendar” to pick a date within the next 30 days. The program had 20 hours of video, although no participant saw every section. The intent-to-treat analysis showed a significant difference from the treatment group at 12.3% vs the controls at 5.0% (OR=2.66, 95% CI, 1.18–5.99).

TABLE 1
Web-based support helps smokers quit

Text messages work

Text messaging may also have a place in supporting smoking cessation efforts. An interesting, although short, study¹³ looked at using text messaging to target younger smokers in New Zealand. This RCT (SOR: B) involved 1705 smokers who had cell phones with text messaging. Researchers sent participants up to 5 messages daily around their quit date, drawing from over 100 messages that could be personalized with individual names/nicknames. The quit rate was doubled 6 weeks out (28% vs 13%; relative risk=2.20; 95% CI, 1.79–2.70).

Rx in hand, support in place

When Ann left my office, she took with her a prescription for varenicline, the state’s quitline number, and the URL for an online support program. Ann was eager to try varenicline: a coworker of hers was using it and doing well. Ann had tried the nicotine patch in the past, but reported that it gave her nightmares. She’d also kept smoking while wearing it. This time, she hoped she’d finally be able to quit for good.

Weighing the drug treatment options

The AHRQ guidelines recommend several types of pharmacotherapy. First-line therapies include different forms of NRT and sustained-release bupropion (Zyban).^2,3

Nicotine replacement therapy doubles the chances of quitting

With NRT, the nicotine in cigarettes is replaced with nicotine from another source. The thought is that by reducing the withdrawal symptoms, the patient is less likely to relapse and resume smoking. Nicotine replacement is available in several forms: gum, transdermal patches, intranasal spray, inhaler, and lozenges.

A Cochrane meta-analysis of NRT¹⁴ (SOR: A) analyzed 123 studies that followed patients for at least 6 months from their quit date. The authors concluded that NRT could almost double a patient’s chances of quitting smoking. The data from various types of NRT revealed the types to be similarly efficacious (TABLE 2). In the treated groups, 17% were abstinent and only 10% were abstinent in the control groups at the various endpoints of the trials. Smokers with higher levels of nicotine dependence as indicated by smoking 10 or more cigarettes daily have higher quit rates using replacement nicotine. Generally, treatments of 8 weeks are as effective as longer courses.

The Cochrane meta-analysis also revealed that:

• Duration of therapy ranges from 3 weeks to 12 months with the various forms of NRT.
• There was no benefit to tapering off the NRT as compared to an abrupt withdrawal.
• Patients are much more likely to relapse after NRT in the first 3 months.
• Combining several forms of NRT may aid a relapsed smoker in another quit attempt. However, the re-attempt should be delayed by a few months, as back-to-back courses are unlikely to improve quit rates.

TABLE 2
Nicotine replacement therapy: Methods are similarly efficacious¹¹

Sustained-release bupropion: Similar results to NRT

The other first-line therapy suggested by the AHRQ guidelines is sustained-release bupropion.^2,3 A separate Cochrane Review¹⁵ analyzed the data from 36 studies using antidepressants and revealed that two thirds of the studies in this meta-analysis used bupropion. The odds of quitting smoking essentially doubled in the placebo-controlled studies. This is a similar effect as NRT. Neither the AHRQ guidelines nor the Cochrane Review recommend bupropion over NRT or vice-versa.

According to the Cochrane Review, there was no benefit to increasing the dose of bupropion from 150 mg to 300 mg daily. Although the initial multi-dose study of bupropion¹⁶ showed a difference, it was not clinically significant by the end of the study. A larger, open-label randomized trial of 1524 smokers¹⁷ followed for 1 year also showed similar results. At the 3-month evaluation, the higher dose had superior efficacy, but that effect was not statistically significant by the end of the study. Lastly, there is no benefit to continuing the bupropion beyond 7 weeks after the target quit date.

With other antidepressants, results vary

The Cochrane Review also looked at other antidepressants. There were 4 RCTs of nortriptyline (Aventyl/Pamelor) without NRT, totaling 777 smokers followed for at least 6 months.^18-21 The pooled data essentially doubled the odds of quitting smoking from 7.0% for the controls to 17.2% in the treated groups (OR=2.79; 95% CI, 1.70–4.59). Adding nortriptyline to NRT did increase the quit rates, but not significantly. The dose used in these studies, at 75 to 150 mg is much lower than that used for depression, where significant side effects often interfere with treatment. Generally the starting dose is 25 mg at bedtime. After 1 week, the dose is increased to 50 mg and the following week, it is increased again to 75 mg. Once on the 75 mg dose for a week, the dose is titrated up only if needed. The titration continues at an additional 25 mg weekly.

One of the 4 placebo-controlled studies²⁰ included an arm of bupropion, producing a head-to-head assessment with nortriptyline (SOR: A). The abstinence rates as indicated by no smoking during the final week of treatment were comparable for the 2 groups receiving active medication. Treatment with bupropion or nortriptyline was significantly more efficacious than placebo. However, the effect was lost at the 1 year continuous abstinence mark; the 2 drugs did not differ from each other or placebo (TABLE 3).

Other antidepressants were evaluated in the Cochrane study.¹⁵ The tricyclic antidepressants doxepin and imipramine (Tofranil) had no long-term studies and neither showed statistically significant differences in smaller trials. Of the selective serotonin reuptake inhibitors (SSRIs), only fluoxetine (Prozac) had any long-term studies, and none noted statistically significant differences. Likewise, venlafaxine (Effexor) had only 1 trial in which the confidence interval did allow for a potentially useful clinical effect, but failed to show a statistically significant increase in 12-month quit rates.

Clonidine is an option, but side effects are an issue

Another Cochrane Review²² looked at the effectiveness of clonidine (Catapres) on smoking cessation. Most of the clonidine studies assessed withdrawal symptoms rather than abstinence. Of those that did assess quit rates, the pooled OR for clonidine compares favorably at 1.89 (95% CI, 1.30–2.74). Unfortunately, clonidine has significant side effects: sedation and postural hypotension. The starting dose is 0.1 mg twice daily, and it may be titrated up to a maximum dose of 0.4 mg daily. It should be used for 3 to 4 weeks only to decrease the symptoms of withdrawal. The smoker should then be weaned off the clonidine.

The anxiolytics were the subject of another Cochrane Review.²³ This review, however, did not recommend any anxiolytics, including diazepam and buspirone, for smoking cessation.

A new category of therapy: Nicotinic receptor agonists

With the US Food and Drug Administration’s approval of varenicline (Chantix) in May 2006, a new class of drugs became available for treatment of tobacco dependence. This α4β2 nicotinic acetylcholine receptor partial agonist was designed as a smoking cessation drug. By releasing dopamine in the brain like nicotine, it prevents craving. However, it also blocks nicotine from binding, thereby preventing the reinforcing effect of continued smoking.

Two RCTs have assessed varenicline against both bupropion and placebo (TABLE 3). Jorenby²⁴ (SOR: A) showed the varenicline-treated participants were significantly more likely to be continuously abstinent at 52 weeks than the placebo-or bupropion-treated groups (23% vs 10.3% placebo [OR=2.66; 95% CI, 1.72–4.11; P<.001] and 14.6% bupropion [OR=1.77; 95% CI, 1.19–2.63; P=.004]). Gonzales²⁵ (SOR: A) likewise showed the varenicline treated smokers were more likely to be continuously abstinent at 52 weeks than the placebo group (21.9% vs 8.4% [OR=3.09; 95% CI, 1.95–4.91; P<.001]). However, the difference between varenicline and bupropion did not reach statistical significance (21.9% vs 16.1% [OR=1.46; 95% CI, 0.99–2.17; P=.057]).

As with other medications, varenicline should be started at a low dose. The patient begins with 0.5 mg nightly for the first 3 nights, then increases to 0.5 mg twice a day for 4 days. The second week, the patient begins the 1 mg twice-daily dosing that is continued through treatment.

TABLE 3
Varenicline, nortriptyline, bupropion—strong allies in patients’ efforts to quit

Vaccines hold the promise of continued abstinence

Several promising ideas for the treatment of tobacco dependence are in development. There are several vaccines being studied.²⁶ When the immune system produces antibodies to nicotine in response to the vaccine, and when these antibodies bind to the nicotine, the resultant compound is too large to cross the blood-brain barrier. This prevents the reinforcing effect of nicotine. Initial studies of vaccines show that smokers do decrease the amount they smoke, but more importantly, abstinence is easier to maintain. However, the vaccine requires frequent boosters to maintain antibody titers that are effective.

NicVAX from Nabi Biopharmaceuticals was placed on a fast track for approval by the Food and Drug Administration. It is, however, still at least a year away from approval. The other 2 nicotine vaccines are probably several years beyond that for approval.²⁷

Researchers are also studying other compounds that block the euphoria associated with smoking.²⁸ The initial studies of rimonabant (Acomplia), a cannabinoid blocker, have shown it is no better than other treatments already available. With its indication in some European countries for weight loss, it offered promise as an important option for patients who are concerned about the weight gain associated with smoking cessation. However, the FDA did not approve rimonabant for tobacco cessation when issuing its initial approval letter for weight loss in 2006. Because of safety concerns, the manufacturer subsequently withdrew the new drug application for rimonabant in 2007.

With much work, our patient kicks the habit

Ann began taking varenicline the day she left the office, and reached her quit date a week later.

At her 1-month follow-up, Ann reported that it was actually easy for her to stay off the cigarettes. With the varenicline, she had lost the desire to smoke. I reminded her to work on the triggers for her smoking: I urged her to make sure that she did not light up when she made her morning coffee or got in the car. I also suggested she put $4 each morning into a jar on her dresser; so she would see how much she saved now that she wasn’t buying cigarettes.

At Ann’s next annual exam, we marked her in the computer system as a reformed smoker. She was very proud of that label. I asked her what she was doing with all that extra cash. She laughed: “My daughter spends it all! but not on cigarettes!”

Acknowledgments

This research was supported by the Intramural Research Program of the NIH, National Institute on Drug Abuse.

Correspondence
Agnes O. Coffay, MD, NIDA/IRP, 5500 Nathan Shock Drive, B altimore, MD 21224-6823. [email protected]

Ann G. is a 34-year-old mother of 2 who had been coming to the office for her annual Pap smear for several years. Her medical history is significant only for her vaginal deliveries and mild GERD. Her medications include oral contraceptive pills (OCPs) and over-the-counter Zantac as needed. on her most recent annual visit, my medical assistant, Tammy, took Ann’s vital signs. The chart had a section about smoking status, and Tammy noted that Ann smoked.

During the office visit, I explained to Ann that her smoking was a serious health risk, and that she needed to quit. She would also need to find a new form of birth control next year, as smoking increases the risks of using OCPs. She nervously laughed off the warning.

The following year, Anne confessed to Tammy that she was still a smoker. Tammy asked her again about quitting. Ann was still adamant: “No way—I can’t do it.” Nonetheless, during the office visit, I brought up the subject of her smoking. She admitted that she was afraid that quitting smoking would cause her to gain weight. I attempted to address her fears, and then talked about other birth control methods to consider. I gave her a 3-month prescription of OCPs, and told her in 3 months we would discuss what she wanted to do about birth control.

Ann faces an uphill battle. The amount of nicotine in cigarettes is increasing,¹ making it harder for her to quit. The good news is that the treatment of tobacco addiction is constantly improving and the number of tools in our arsenal is growing. In fact, there are many resources that we can try before turning to the prescription pad. However, when needed, pharmacotherapy is an important adjunct for achieving abstinence.

“5-A” strategy sets stage for success

The Agency for Healthcare Research and Quality (AHRQ) has published Treating Tobacco Use and Dependence, a useful guide for helping patients quit.^2,3 These guidelines discuss many aspects of tobacco cessation, from counseling to pharmacotherapy to reimbursement issues. The guidelines break down the smoking cessation process into the 5 A’s:

1. Ask each patient about her smoking status.
2. Advise each patient who smokes that she needs to stop smoking.
3. Assess your patient’s willingness to make a quit attempt in the next 30 days.
4. Assist your patient either in making this quit attempt or in motivating her to consider a quit attempt later.
5. Arrange close follow-up of any quit attempts to help prevent relapse.

The Ask and Act program from the American Academy of Family Physicians (AAFP) outlines a similar strategy.⁴ The program instructs physicians to Ask every patient about her tobacco use and to Act to help her quit, via on-or off-site counseling, quitlines, patient education materials, self-help guides or Web sites, cessation classes, and pharmacotherapy.

Take advantage of every opportunity you have to discuss the issue with patients; short conversations can make a difference. A Cochrane Review of 39 trials including 31,000 smokers⁵ revealed that even brief advice—simply encouraging patients to quit—was statistically significant (odds ratio [OR]=1.74; 95% confidence interval [CI], 1.48–2.05). The pooled data generated a quit rate difference of 2.5%: for every 40 people who were told to quit, 1 more smoker would.

Empower the office staff

Enlisting the help of the office staff can have a significant impact on the health of the patients. A proactive approach was studied by Fiore et al.⁶ Medical assistants, while assessing smoking status, invited all smokers to participate in a cessation study. (The assistants received periodic thank-you gifts for their efforts.)

The participants were randomized to either self-selected treatment or nicotine replacement therapy (NRT) patches, with or without a support program. Some who received the patches and support program also received individual counseling. Fiore et al showed that the majority of smokers were open to attempts to quit smoking. The 13% point-prevalence abstinence rate 1 year out is comparable with the rate obtained (14%) with smokers volunteering for NRT studies in the Cochrane review of 39 trials, noted earlier.

Likewise, in a randomized controlled trial (RCT) involving community-based primary care clinics, Katz⁷ demonstrated that intake clinicians could also play an important role in smoking cessation (SOR: A). In the study, researchers trained intake clinicians (including registered nurses, license practical nurses, and medical assistants) to identify smokers, provide brief counseling, and assist in their preparation to quit. Patients were offered vouchers for patches and a counselor’s business card. Intake clinicians received periodic feedback on their performance based on exit interviews of the patients. The researchers found that these interventions had a statistically significant effect in moderate-to-heavy smokers in quit attempts, quit rates, and continuous abstinence.

Our patient has a change of heart

At the 3-month follow up, Tammy learned that Ann was still smoking—but she now wanted to quit. Ann said that she’d found a pack of cigarettes in her 14-year-old daughter’s backpack, and felt that the only way to prevent her from getting hooked was to set a good example.

Tammy gave her the state’s quitline number and suggested some online quitting programs. Tammy worked with Ann to choose her target quit date and to pick the Web-based program she was going to use. Ann said that she liked the idea that she could go online whenever she needed support. She also liked the fact that she could put her quit date into the system, so it would give her timely reminders of all her reasons to quit when she logged on.

I wrote Ann a prescription for varenicline and her OCPs, and told her I wanted to see her in 4 weeks. For her part, Tammy added Ann to her list of patients to call the day after her quit date. Tammy makes this her practice with patients because she knows that one well-timed phone call can be the key to a successful quitting attempt.

Outside support improves abstinence rates

Improving your patients’ chances of long-term abstinence hinges, in part, on making the most of outside support. In many cases, your patients can take advantage of them without leaving their homes.

Outside support improves abstinence rates

Quitlines increase quit rates, decrease relapse Telephone counseling is an effective support system.⁸ Smokers who call to a single number (800-QUITNOW)—a service provided by the National Cancer Institute—are directed to the quitline for their state. Also, smokers can call the National Cancer Institute directly at their quitline (877-44U-QUIT). Calling a quitline provides smokers with real-time counseling and information about how to quit smoking. Quitlines can be appealing to those patients who are uncomfortable discussing their smoking in a group—and it’s free to the patient.

The research supports the use of such help lines. Zhu’s study⁹ of the California Smokers’ Helpline (SOR: A) was a proactive protocol where smokers were funneled into a research trial when the help line was overwhelmed. The smokers in the treatment arm of this RCT were assigned a counselor who called the smokers as many as 6 times, following a relapse-sensitive schedule. The 12-month abstinence rate increased from 4.1% to 7.5% (P<.001) in the group that had close telephone contact. This improved quit rate reflected both an increase of percentage of smokers who quit and, more importantly, a decrease in quitters who relapsed.

Another prospective RCT¹⁰ (SOR: A) enrolled patients from Veterans Affairs (VA) medical centers and involved the same proactive telephone protocol as Zhu used. The treatment group was offered telephone counseling as well as pharmacotherapy; the control group had access to the regular smoking cessation program of the VA system. Regardless of which group an individual was assigned, if that participant used both the counseling and the pharmacotherapy, the quit rate was similar: control (12.7%) and treatment (11.9%). However, only 18% of the controls used both services. The treatment group accessed the combined programs of counseling and medications at a rate of 88%. This led to the difference in 6-month abstinence rates of 13.0% in the treatment group and 4.1% in the control group (OR=3.50; 95% CI, 1.99–6.15). Patients who were directed to and enrolled in treatment programs were therefore more likely to attempt to quit and remain abstinent for up to 6 months.

Web-based programs offer reminders

Like quitlines, Web-based programs offer smokers immediate feedback to help them quit. Many of the programs include links to quitting resources, stories from former smokers and cancer patients, live advice from counselors, and message boards (TABLE 1). Web-based programs have been shown to help improve quit rates.

One study¹¹ compared 2 Web programs involving 11,969 smokers. This RCT (SOR: B) looked at an interactive program based, in part, on the AHRQ treatment guidelines. This program generates personalized letters for the participants along with monthly e-mail reminders. A modified program was used as the control. The control program was developed by a maker of NRT products, and contained more information about nicotine than about tobacco dependence and cessation. This program was also shorter than the interactive program, which was designed to assist smoking cessation.

Both programs improved quit rates: 10.9% for the interactive program and 8% for the modified/control program, compared with 3.3% for no treatment at all. Although this study was based on participant reports of abstinence over the previous 7 days, and had low follow-up rates (which Internet studies tend to have), the interactive program produced 1 more quitter for every 26 participants than the modified (control) program, using an intent-to-treat analysis (14.6% vs. 10.7%, P<.001, OR=1.43, 95% CI, 1.28–1.59).

Another study¹² looked at the use of a more extensive Web site, combining video, audio, and text. This RCT (SOR: B) was fully automated and delivered entirely by computer. Again, using the AHRQ guidelines and other sources, researchers designed a series of 5 modules to simulate working with a live counselor. There were 13 different versions, to match the demographics of the participant. The modules ended with a “quit calendar” to pick a date within the next 30 days. The program had 20 hours of video, although no participant saw every section. The intent-to-treat analysis showed a significant difference from the treatment group at 12.3% vs the controls at 5.0% (OR=2.66, 95% CI, 1.18–5.99).

TABLE 1
Web-based support helps smokers quit

Text messages work

Text messaging may also have a place in supporting smoking cessation efforts. An interesting, although short, study¹³ looked at using text messaging to target younger smokers in New Zealand. This RCT (SOR: B) involved 1705 smokers who had cell phones with text messaging. Researchers sent participants up to 5 messages daily around their quit date, drawing from over 100 messages that could be personalized with individual names/nicknames. The quit rate was doubled 6 weeks out (28% vs 13%; relative risk=2.20; 95% CI, 1.79–2.70).

Rx in hand, support in place

When Ann left my office, she took with her a prescription for varenicline, the state’s quitline number, and the URL for an online support program. Ann was eager to try varenicline: a coworker of hers was using it and doing well. Ann had tried the nicotine patch in the past, but reported that it gave her nightmares. She’d also kept smoking while wearing it. This time, she hoped she’d finally be able to quit for good.

Weighing the drug treatment options

The AHRQ guidelines recommend several types of pharmacotherapy. First-line therapies include different forms of NRT and sustained-release bupropion (Zyban).^2,3

Nicotine replacement therapy doubles the chances of quitting

With NRT, the nicotine in cigarettes is replaced with nicotine from another source. The thought is that by reducing the withdrawal symptoms, the patient is less likely to relapse and resume smoking. Nicotine replacement is available in several forms: gum, transdermal patches, intranasal spray, inhaler, and lozenges.

A Cochrane meta-analysis of NRT¹⁴ (SOR: A) analyzed 123 studies that followed patients for at least 6 months from their quit date. The authors concluded that NRT could almost double a patient’s chances of quitting smoking. The data from various types of NRT revealed the types to be similarly efficacious (TABLE 2). In the treated groups, 17% were abstinent and only 10% were abstinent in the control groups at the various endpoints of the trials. Smokers with higher levels of nicotine dependence as indicated by smoking 10 or more cigarettes daily have higher quit rates using replacement nicotine. Generally, treatments of 8 weeks are as effective as longer courses.

The Cochrane meta-analysis also revealed that:

• Duration of therapy ranges from 3 weeks to 12 months with the various forms of NRT.
• There was no benefit to tapering off the NRT as compared to an abrupt withdrawal.
• Patients are much more likely to relapse after NRT in the first 3 months.
• Combining several forms of NRT may aid a relapsed smoker in another quit attempt. However, the re-attempt should be delayed by a few months, as back-to-back courses are unlikely to improve quit rates.

TABLE 2
Nicotine replacement therapy: Methods are similarly efficacious¹¹

Sustained-release bupropion: Similar results to NRT

The other first-line therapy suggested by the AHRQ guidelines is sustained-release bupropion.^2,3 A separate Cochrane Review¹⁵ analyzed the data from 36 studies using antidepressants and revealed that two thirds of the studies in this meta-analysis used bupropion. The odds of quitting smoking essentially doubled in the placebo-controlled studies. This is a similar effect as NRT. Neither the AHRQ guidelines nor the Cochrane Review recommend bupropion over NRT or vice-versa.

According to the Cochrane Review, there was no benefit to increasing the dose of bupropion from 150 mg to 300 mg daily. Although the initial multi-dose study of bupropion¹⁶ showed a difference, it was not clinically significant by the end of the study. A larger, open-label randomized trial of 1524 smokers¹⁷ followed for 1 year also showed similar results. At the 3-month evaluation, the higher dose had superior efficacy, but that effect was not statistically significant by the end of the study. Lastly, there is no benefit to continuing the bupropion beyond 7 weeks after the target quit date.

With other antidepressants, results vary

The Cochrane Review also looked at other antidepressants. There were 4 RCTs of nortriptyline (Aventyl/Pamelor) without NRT, totaling 777 smokers followed for at least 6 months.^18-21 The pooled data essentially doubled the odds of quitting smoking from 7.0% for the controls to 17.2% in the treated groups (OR=2.79; 95% CI, 1.70–4.59). Adding nortriptyline to NRT did increase the quit rates, but not significantly. The dose used in these studies, at 75 to 150 mg is much lower than that used for depression, where significant side effects often interfere with treatment. Generally the starting dose is 25 mg at bedtime. After 1 week, the dose is increased to 50 mg and the following week, it is increased again to 75 mg. Once on the 75 mg dose for a week, the dose is titrated up only if needed. The titration continues at an additional 25 mg weekly.

One of the 4 placebo-controlled studies²⁰ included an arm of bupropion, producing a head-to-head assessment with nortriptyline (SOR: A). The abstinence rates as indicated by no smoking during the final week of treatment were comparable for the 2 groups receiving active medication. Treatment with bupropion or nortriptyline was significantly more efficacious than placebo. However, the effect was lost at the 1 year continuous abstinence mark; the 2 drugs did not differ from each other or placebo (TABLE 3).

Other antidepressants were evaluated in the Cochrane study.¹⁵ The tricyclic antidepressants doxepin and imipramine (Tofranil) had no long-term studies and neither showed statistically significant differences in smaller trials. Of the selective serotonin reuptake inhibitors (SSRIs), only fluoxetine (Prozac) had any long-term studies, and none noted statistically significant differences. Likewise, venlafaxine (Effexor) had only 1 trial in which the confidence interval did allow for a potentially useful clinical effect, but failed to show a statistically significant increase in 12-month quit rates.

Clonidine is an option, but side effects are an issue

Another Cochrane Review²² looked at the effectiveness of clonidine (Catapres) on smoking cessation. Most of the clonidine studies assessed withdrawal symptoms rather than abstinence. Of those that did assess quit rates, the pooled OR for clonidine compares favorably at 1.89 (95% CI, 1.30–2.74). Unfortunately, clonidine has significant side effects: sedation and postural hypotension. The starting dose is 0.1 mg twice daily, and it may be titrated up to a maximum dose of 0.4 mg daily. It should be used for 3 to 4 weeks only to decrease the symptoms of withdrawal. The smoker should then be weaned off the clonidine.

The anxiolytics were the subject of another Cochrane Review.²³ This review, however, did not recommend any anxiolytics, including diazepam and buspirone, for smoking cessation.

A new category of therapy: Nicotinic receptor agonists

With the US Food and Drug Administration’s approval of varenicline (Chantix) in May 2006, a new class of drugs became available for treatment of tobacco dependence. This α4β2 nicotinic acetylcholine receptor partial agonist was designed as a smoking cessation drug. By releasing dopamine in the brain like nicotine, it prevents craving. However, it also blocks nicotine from binding, thereby preventing the reinforcing effect of continued smoking.

Two RCTs have assessed varenicline against both bupropion and placebo (TABLE 3). Jorenby²⁴ (SOR: A) showed the varenicline-treated participants were significantly more likely to be continuously abstinent at 52 weeks than the placebo-or bupropion-treated groups (23% vs 10.3% placebo [OR=2.66; 95% CI, 1.72–4.11; P<.001] and 14.6% bupropion [OR=1.77; 95% CI, 1.19–2.63; P=.004]). Gonzales²⁵ (SOR: A) likewise showed the varenicline treated smokers were more likely to be continuously abstinent at 52 weeks than the placebo group (21.9% vs 8.4% [OR=3.09; 95% CI, 1.95–4.91; P<.001]). However, the difference between varenicline and bupropion did not reach statistical significance (21.9% vs 16.1% [OR=1.46; 95% CI, 0.99–2.17; P=.057]).

As with other medications, varenicline should be started at a low dose. The patient begins with 0.5 mg nightly for the first 3 nights, then increases to 0.5 mg twice a day for 4 days. The second week, the patient begins the 1 mg twice-daily dosing that is continued through treatment.

TABLE 3
Varenicline, nortriptyline, bupropion—strong allies in patients’ efforts to quit

Vaccines hold the promise of continued abstinence

Several promising ideas for the treatment of tobacco dependence are in development. There are several vaccines being studied.²⁶ When the immune system produces antibodies to nicotine in response to the vaccine, and when these antibodies bind to the nicotine, the resultant compound is too large to cross the blood-brain barrier. This prevents the reinforcing effect of nicotine. Initial studies of vaccines show that smokers do decrease the amount they smoke, but more importantly, abstinence is easier to maintain. However, the vaccine requires frequent boosters to maintain antibody titers that are effective.

NicVAX from Nabi Biopharmaceuticals was placed on a fast track for approval by the Food and Drug Administration. It is, however, still at least a year away from approval. The other 2 nicotine vaccines are probably several years beyond that for approval.²⁷

Researchers are also studying other compounds that block the euphoria associated with smoking.²⁸ The initial studies of rimonabant (Acomplia), a cannabinoid blocker, have shown it is no better than other treatments already available. With its indication in some European countries for weight loss, it offered promise as an important option for patients who are concerned about the weight gain associated with smoking cessation. However, the FDA did not approve rimonabant for tobacco cessation when issuing its initial approval letter for weight loss in 2006. Because of safety concerns, the manufacturer subsequently withdrew the new drug application for rimonabant in 2007.

With much work, our patient kicks the habit

Ann began taking varenicline the day she left the office, and reached her quit date a week later.

At her 1-month follow-up, Ann reported that it was actually easy for her to stay off the cigarettes. With the varenicline, she had lost the desire to smoke. I reminded her to work on the triggers for her smoking: I urged her to make sure that she did not light up when she made her morning coffee or got in the car. I also suggested she put $4 each morning into a jar on her dresser; so she would see how much she saved now that she wasn’t buying cigarettes.

At Ann’s next annual exam, we marked her in the computer system as a reformed smoker. She was very proud of that label. I asked her what she was doing with all that extra cash. She laughed: “My daughter spends it all! but not on cigarettes!”

Acknowledgments

This research was supported by the Intramural Research Program of the NIH, National Institute on Drug Abuse.

Correspondence
Agnes O. Coffay, MD, NIDA/IRP, 5500 Nathan Shock Drive, B altimore, MD 21224-6823. [email protected]

Ann G. is a 34-year-old mother of 2 who had been coming to the office for her annual Pap smear for several years. Her medical history is significant only for her vaginal deliveries and mild GERD. Her medications include oral contraceptive pills (OCPs) and over-the-counter Zantac as needed. on her most recent annual visit, my medical assistant, Tammy, took Ann’s vital signs. The chart had a section about smoking status, and Tammy noted that Ann smoked.

During the office visit, I explained to Ann that her smoking was a serious health risk, and that she needed to quit. She would also need to find a new form of birth control next year, as smoking increases the risks of using OCPs. She nervously laughed off the warning.

The following year, Anne confessed to Tammy that she was still a smoker. Tammy asked her again about quitting. Ann was still adamant: “No way—I can’t do it.” Nonetheless, during the office visit, I brought up the subject of her smoking. She admitted that she was afraid that quitting smoking would cause her to gain weight. I attempted to address her fears, and then talked about other birth control methods to consider. I gave her a 3-month prescription of OCPs, and told her in 3 months we would discuss what she wanted to do about birth control.

Ann faces an uphill battle. The amount of nicotine in cigarettes is increasing,¹ making it harder for her to quit. The good news is that the treatment of tobacco addiction is constantly improving and the number of tools in our arsenal is growing. In fact, there are many resources that we can try before turning to the prescription pad. However, when needed, pharmacotherapy is an important adjunct for achieving abstinence.

“5-A” strategy sets stage for success

The Agency for Healthcare Research and Quality (AHRQ) has published Treating Tobacco Use and Dependence, a useful guide for helping patients quit.^2,3 These guidelines discuss many aspects of tobacco cessation, from counseling to pharmacotherapy to reimbursement issues. The guidelines break down the smoking cessation process into the 5 A’s:

1. Ask each patient about her smoking status.
2. Advise each patient who smokes that she needs to stop smoking.
3. Assess your patient’s willingness to make a quit attempt in the next 30 days.
4. Assist your patient either in making this quit attempt or in motivating her to consider a quit attempt later.
5. Arrange close follow-up of any quit attempts to help prevent relapse.

The Ask and Act program from the American Academy of Family Physicians (AAFP) outlines a similar strategy.⁴ The program instructs physicians to Ask every patient about her tobacco use and to Act to help her quit, via on-or off-site counseling, quitlines, patient education materials, self-help guides or Web sites, cessation classes, and pharmacotherapy.

Take advantage of every opportunity you have to discuss the issue with patients; short conversations can make a difference. A Cochrane Review of 39 trials including 31,000 smokers⁵ revealed that even brief advice—simply encouraging patients to quit—was statistically significant (odds ratio [OR]=1.74; 95% confidence interval [CI], 1.48–2.05). The pooled data generated a quit rate difference of 2.5%: for every 40 people who were told to quit, 1 more smoker would.

Empower the office staff

Enlisting the help of the office staff can have a significant impact on the health of the patients. A proactive approach was studied by Fiore et al.⁶ Medical assistants, while assessing smoking status, invited all smokers to participate in a cessation study. (The assistants received periodic thank-you gifts for their efforts.)

The participants were randomized to either self-selected treatment or nicotine replacement therapy (NRT) patches, with or without a support program. Some who received the patches and support program also received individual counseling. Fiore et al showed that the majority of smokers were open to attempts to quit smoking. The 13% point-prevalence abstinence rate 1 year out is comparable with the rate obtained (14%) with smokers volunteering for NRT studies in the Cochrane review of 39 trials, noted earlier.

Likewise, in a randomized controlled trial (RCT) involving community-based primary care clinics, Katz⁷ demonstrated that intake clinicians could also play an important role in smoking cessation (SOR: A). In the study, researchers trained intake clinicians (including registered nurses, license practical nurses, and medical assistants) to identify smokers, provide brief counseling, and assist in their preparation to quit. Patients were offered vouchers for patches and a counselor’s business card. Intake clinicians received periodic feedback on their performance based on exit interviews of the patients. The researchers found that these interventions had a statistically significant effect in moderate-to-heavy smokers in quit attempts, quit rates, and continuous abstinence.

Our patient has a change of heart

At the 3-month follow up, Tammy learned that Ann was still smoking—but she now wanted to quit. Ann said that she’d found a pack of cigarettes in her 14-year-old daughter’s backpack, and felt that the only way to prevent her from getting hooked was to set a good example.

Tammy gave her the state’s quitline number and suggested some online quitting programs. Tammy worked with Ann to choose her target quit date and to pick the Web-based program she was going to use. Ann said that she liked the idea that she could go online whenever she needed support. She also liked the fact that she could put her quit date into the system, so it would give her timely reminders of all her reasons to quit when she logged on.

I wrote Ann a prescription for varenicline and her OCPs, and told her I wanted to see her in 4 weeks. For her part, Tammy added Ann to her list of patients to call the day after her quit date. Tammy makes this her practice with patients because she knows that one well-timed phone call can be the key to a successful quitting attempt.

Outside support improves abstinence rates

Improving your patients’ chances of long-term abstinence hinges, in part, on making the most of outside support. In many cases, your patients can take advantage of them without leaving their homes.

Outside support improves abstinence rates

Quitlines increase quit rates, decrease relapse Telephone counseling is an effective support system.⁸ Smokers who call to a single number (800-QUITNOW)—a service provided by the National Cancer Institute—are directed to the quitline for their state. Also, smokers can call the National Cancer Institute directly at their quitline (877-44U-QUIT). Calling a quitline provides smokers with real-time counseling and information about how to quit smoking. Quitlines can be appealing to those patients who are uncomfortable discussing their smoking in a group—and it’s free to the patient.

The research supports the use of such help lines. Zhu’s study⁹ of the California Smokers’ Helpline (SOR: A) was a proactive protocol where smokers were funneled into a research trial when the help line was overwhelmed. The smokers in the treatment arm of this RCT were assigned a counselor who called the smokers as many as 6 times, following a relapse-sensitive schedule. The 12-month abstinence rate increased from 4.1% to 7.5% (P<.001) in the group that had close telephone contact. This improved quit rate reflected both an increase of percentage of smokers who quit and, more importantly, a decrease in quitters who relapsed.

Another prospective RCT¹⁰ (SOR: A) enrolled patients from Veterans Affairs (VA) medical centers and involved the same proactive telephone protocol as Zhu used. The treatment group was offered telephone counseling as well as pharmacotherapy; the control group had access to the regular smoking cessation program of the VA system. Regardless of which group an individual was assigned, if that participant used both the counseling and the pharmacotherapy, the quit rate was similar: control (12.7%) and treatment (11.9%). However, only 18% of the controls used both services. The treatment group accessed the combined programs of counseling and medications at a rate of 88%. This led to the difference in 6-month abstinence rates of 13.0% in the treatment group and 4.1% in the control group (OR=3.50; 95% CI, 1.99–6.15). Patients who were directed to and enrolled in treatment programs were therefore more likely to attempt to quit and remain abstinent for up to 6 months.

Web-based programs offer reminders

Like quitlines, Web-based programs offer smokers immediate feedback to help them quit. Many of the programs include links to quitting resources, stories from former smokers and cancer patients, live advice from counselors, and message boards (TABLE 1). Web-based programs have been shown to help improve quit rates.

One study¹¹ compared 2 Web programs involving 11,969 smokers. This RCT (SOR: B) looked at an interactive program based, in part, on the AHRQ treatment guidelines. This program generates personalized letters for the participants along with monthly e-mail reminders. A modified program was used as the control. The control program was developed by a maker of NRT products, and contained more information about nicotine than about tobacco dependence and cessation. This program was also shorter than the interactive program, which was designed to assist smoking cessation.

Both programs improved quit rates: 10.9% for the interactive program and 8% for the modified/control program, compared with 3.3% for no treatment at all. Although this study was based on participant reports of abstinence over the previous 7 days, and had low follow-up rates (which Internet studies tend to have), the interactive program produced 1 more quitter for every 26 participants than the modified (control) program, using an intent-to-treat analysis (14.6% vs. 10.7%, P<.001, OR=1.43, 95% CI, 1.28–1.59).

Another study¹² looked at the use of a more extensive Web site, combining video, audio, and text. This RCT (SOR: B) was fully automated and delivered entirely by computer. Again, using the AHRQ guidelines and other sources, researchers designed a series of 5 modules to simulate working with a live counselor. There were 13 different versions, to match the demographics of the participant. The modules ended with a “quit calendar” to pick a date within the next 30 days. The program had 20 hours of video, although no participant saw every section. The intent-to-treat analysis showed a significant difference from the treatment group at 12.3% vs the controls at 5.0% (OR=2.66, 95% CI, 1.18–5.99).

TABLE 1
Web-based support helps smokers quit

Text messages work

Text messaging may also have a place in supporting smoking cessation efforts. An interesting, although short, study¹³ looked at using text messaging to target younger smokers in New Zealand. This RCT (SOR: B) involved 1705 smokers who had cell phones with text messaging. Researchers sent participants up to 5 messages daily around their quit date, drawing from over 100 messages that could be personalized with individual names/nicknames. The quit rate was doubled 6 weeks out (28% vs 13%; relative risk=2.20; 95% CI, 1.79–2.70).

Rx in hand, support in place

When Ann left my office, she took with her a prescription for varenicline, the state’s quitline number, and the URL for an online support program. Ann was eager to try varenicline: a coworker of hers was using it and doing well. Ann had tried the nicotine patch in the past, but reported that it gave her nightmares. She’d also kept smoking while wearing it. This time, she hoped she’d finally be able to quit for good.

Weighing the drug treatment options

The AHRQ guidelines recommend several types of pharmacotherapy. First-line therapies include different forms of NRT and sustained-release bupropion (Zyban).^2,3

Nicotine replacement therapy doubles the chances of quitting

With NRT, the nicotine in cigarettes is replaced with nicotine from another source. The thought is that by reducing the withdrawal symptoms, the patient is less likely to relapse and resume smoking. Nicotine replacement is available in several forms: gum, transdermal patches, intranasal spray, inhaler, and lozenges.

A Cochrane meta-analysis of NRT¹⁴ (SOR: A) analyzed 123 studies that followed patients for at least 6 months from their quit date. The authors concluded that NRT could almost double a patient’s chances of quitting smoking. The data from various types of NRT revealed the types to be similarly efficacious (TABLE 2). In the treated groups, 17% were abstinent and only 10% were abstinent in the control groups at the various endpoints of the trials. Smokers with higher levels of nicotine dependence as indicated by smoking 10 or more cigarettes daily have higher quit rates using replacement nicotine. Generally, treatments of 8 weeks are as effective as longer courses.

The Cochrane meta-analysis also revealed that:

• Duration of therapy ranges from 3 weeks to 12 months with the various forms of NRT.
• There was no benefit to tapering off the NRT as compared to an abrupt withdrawal.
• Patients are much more likely to relapse after NRT in the first 3 months.
• Combining several forms of NRT may aid a relapsed smoker in another quit attempt. However, the re-attempt should be delayed by a few months, as back-to-back courses are unlikely to improve quit rates.

TABLE 2
Nicotine replacement therapy: Methods are similarly efficacious¹¹

Sustained-release bupropion: Similar results to NRT

The other first-line therapy suggested by the AHRQ guidelines is sustained-release bupropion.^2,3 A separate Cochrane Review¹⁵ analyzed the data from 36 studies using antidepressants and revealed that two thirds of the studies in this meta-analysis used bupropion. The odds of quitting smoking essentially doubled in the placebo-controlled studies. This is a similar effect as NRT. Neither the AHRQ guidelines nor the Cochrane Review recommend bupropion over NRT or vice-versa.

According to the Cochrane Review, there was no benefit to increasing the dose of bupropion from 150 mg to 300 mg daily. Although the initial multi-dose study of bupropion¹⁶ showed a difference, it was not clinically significant by the end of the study. A larger, open-label randomized trial of 1524 smokers¹⁷ followed for 1 year also showed similar results. At the 3-month evaluation, the higher dose had superior efficacy, but that effect was not statistically significant by the end of the study. Lastly, there is no benefit to continuing the bupropion beyond 7 weeks after the target quit date.

With other antidepressants, results vary

The Cochrane Review also looked at other antidepressants. There were 4 RCTs of nortriptyline (Aventyl/Pamelor) without NRT, totaling 777 smokers followed for at least 6 months.^18-21 The pooled data essentially doubled the odds of quitting smoking from 7.0% for the controls to 17.2% in the treated groups (OR=2.79; 95% CI, 1.70–4.59). Adding nortriptyline to NRT did increase the quit rates, but not significantly. The dose used in these studies, at 75 to 150 mg is much lower than that used for depression, where significant side effects often interfere with treatment. Generally the starting dose is 25 mg at bedtime. After 1 week, the dose is increased to 50 mg and the following week, it is increased again to 75 mg. Once on the 75 mg dose for a week, the dose is titrated up only if needed. The titration continues at an additional 25 mg weekly.

One of the 4 placebo-controlled studies²⁰ included an arm of bupropion, producing a head-to-head assessment with nortriptyline (SOR: A). The abstinence rates as indicated by no smoking during the final week of treatment were comparable for the 2 groups receiving active medication. Treatment with bupropion or nortriptyline was significantly more efficacious than placebo. However, the effect was lost at the 1 year continuous abstinence mark; the 2 drugs did not differ from each other or placebo (TABLE 3).

Other antidepressants were evaluated in the Cochrane study.¹⁵ The tricyclic antidepressants doxepin and imipramine (Tofranil) had no long-term studies and neither showed statistically significant differences in smaller trials. Of the selective serotonin reuptake inhibitors (SSRIs), only fluoxetine (Prozac) had any long-term studies, and none noted statistically significant differences. Likewise, venlafaxine (Effexor) had only 1 trial in which the confidence interval did allow for a potentially useful clinical effect, but failed to show a statistically significant increase in 12-month quit rates.

Clonidine is an option, but side effects are an issue

Another Cochrane Review²² looked at the effectiveness of clonidine (Catapres) on smoking cessation. Most of the clonidine studies assessed withdrawal symptoms rather than abstinence. Of those that did assess quit rates, the pooled OR for clonidine compares favorably at 1.89 (95% CI, 1.30–2.74). Unfortunately, clonidine has significant side effects: sedation and postural hypotension. The starting dose is 0.1 mg twice daily, and it may be titrated up to a maximum dose of 0.4 mg daily. It should be used for 3 to 4 weeks only to decrease the symptoms of withdrawal. The smoker should then be weaned off the clonidine.

The anxiolytics were the subject of another Cochrane Review.²³ This review, however, did not recommend any anxiolytics, including diazepam and buspirone, for smoking cessation.

A new category of therapy: Nicotinic receptor agonists

With the US Food and Drug Administration’s approval of varenicline (Chantix) in May 2006, a new class of drugs became available for treatment of tobacco dependence. This α4β2 nicotinic acetylcholine receptor partial agonist was designed as a smoking cessation drug. By releasing dopamine in the brain like nicotine, it prevents craving. However, it also blocks nicotine from binding, thereby preventing the reinforcing effect of continued smoking.

Two RCTs have assessed varenicline against both bupropion and placebo (TABLE 3). Jorenby²⁴ (SOR: A) showed the varenicline-treated participants were significantly more likely to be continuously abstinent at 52 weeks than the placebo-or bupropion-treated groups (23% vs 10.3% placebo [OR=2.66; 95% CI, 1.72–4.11; P<.001] and 14.6% bupropion [OR=1.77; 95% CI, 1.19–2.63; P=.004]). Gonzales²⁵ (SOR: A) likewise showed the varenicline treated smokers were more likely to be continuously abstinent at 52 weeks than the placebo group (21.9% vs 8.4% [OR=3.09; 95% CI, 1.95–4.91; P<.001]). However, the difference between varenicline and bupropion did not reach statistical significance (21.9% vs 16.1% [OR=1.46; 95% CI, 0.99–2.17; P=.057]).

As with other medications, varenicline should be started at a low dose. The patient begins with 0.5 mg nightly for the first 3 nights, then increases to 0.5 mg twice a day for 4 days. The second week, the patient begins the 1 mg twice-daily dosing that is continued through treatment.

TABLE 3
Varenicline, nortriptyline, bupropion—strong allies in patients’ efforts to quit

Vaccines hold the promise of continued abstinence

Several promising ideas for the treatment of tobacco dependence are in development. There are several vaccines being studied.²⁶ When the immune system produces antibodies to nicotine in response to the vaccine, and when these antibodies bind to the nicotine, the resultant compound is too large to cross the blood-brain barrier. This prevents the reinforcing effect of nicotine. Initial studies of vaccines show that smokers do decrease the amount they smoke, but more importantly, abstinence is easier to maintain. However, the vaccine requires frequent boosters to maintain antibody titers that are effective.

NicVAX from Nabi Biopharmaceuticals was placed on a fast track for approval by the Food and Drug Administration. It is, however, still at least a year away from approval. The other 2 nicotine vaccines are probably several years beyond that for approval.²⁷

Researchers are also studying other compounds that block the euphoria associated with smoking.²⁸ The initial studies of rimonabant (Acomplia), a cannabinoid blocker, have shown it is no better than other treatments already available. With its indication in some European countries for weight loss, it offered promise as an important option for patients who are concerned about the weight gain associated with smoking cessation. However, the FDA did not approve rimonabant for tobacco cessation when issuing its initial approval letter for weight loss in 2006. Because of safety concerns, the manufacturer subsequently withdrew the new drug application for rimonabant in 2007.

With much work, our patient kicks the habit

Ann began taking varenicline the day she left the office, and reached her quit date a week later.

At her 1-month follow-up, Ann reported that it was actually easy for her to stay off the cigarettes. With the varenicline, she had lost the desire to smoke. I reminded her to work on the triggers for her smoking: I urged her to make sure that she did not light up when she made her morning coffee or got in the car. I also suggested she put $4 each morning into a jar on her dresser; so she would see how much she saved now that she wasn’t buying cigarettes.

At Ann’s next annual exam, we marked her in the computer system as a reformed smoker. She was very proud of that label. I asked her what she was doing with all that extra cash. She laughed: “My daughter spends it all! but not on cigarettes!”

Acknowledgments

This research was supported by the Intramural Research Program of the NIH, National Institute on Drug Abuse.

Correspondence
Agnes O. Coffay, MD, NIDA/IRP, 5500 Nathan Shock Drive, B altimore, MD 21224-6823. [email protected]

Baby M is a 12-hour-old 40-week gestation boy, without any risk factors, whom you delivered vaginally to a first-time mother last evening. The following morning, his mother tells you that while she intends to breastfeed exclusively, she was told by a staff member in the nursery to give the baby some formula “until your milk comes in, so he won’t be so cranky.” The child appears entirely well.

The next morning, at 36-hours-of-age, Baby M still appears well, except that he is jaundiced. His bilirubin level is 15—just above the level at which phototherapy should be started, but well below the level at which an exchange transfusion is required. You order a transfer to the nursery for phototherapy.

You order IV fluids to prevent dehydration and to enhance the excretion of bilirubin. To bring his bilirubin level down more quickly, you advise the mother to temporarily stop breastfeeding and feed him a standard infant formula. The nursing staff gives the mother a breast pump to use until breastfeeding is re-started.

Forty-eight hours later Baby M is ready for discharge, and the mother has questions about infant feeding practices. You and the nurse explain to the mother that “breast milk provides complete nutrition” for her baby and that the baby will need no other nutrition until he starts solid foods at around 4 to 6 months of age.

Routine care, but was it appropriate?

The care you provided to Baby M and the advice you offered his mother might seem pretty routine, but look again. Was it appropriate?

In fact, your care—and that of the nursing staff—diverged from the evidence in 4 ways:

1. There was no need to give Baby M formula until his mother’s milk came in (Strength of Recommendation [SOR]: C).
2. It was not necessary to use IV fluids to treat jaundice in this otherwise healthy term newborn (SOR: C). IV fluids do not bring down bilirubin levels, and even with mild dehydration, the best fluid therapy is breast milk or formula.
3. There was no need to temporarily discontinue breastfeeding during phototherapy to bring Baby M’s bilirubin level down more quickly (SOR: C). While doing so may bring levels down a bit faster, it is not worth the risk that the mother may opt to discontinue breastfeeding entirely.
4. You should not have told Baby M’s mother that “breast milk provides complete nutrition.” In fact, the American Academy of Pediatrics (AAP) recommends that all breast-fed newborns receive supplemental vitamin D (SOR: B).

12 common beliefs that require a second look

Medicine has a long history of promoting practices like the ones described here, which are either not based upon scientific evidence or are in direct contradiction to existing evidence. Often, we do not realize how much of our advice is based on misunderstandings passed down from our teachers and absorbed from the cultural milieu in which we were raised.

In this article, we will review the most current evidence regarding 12 commonly held beliefs regarding newborn care. These beliefs relate to breast milk, breastfeeding, pacifier use, emesis, umbilical cord care, and jaundice. We’ve included them because these beliefs give rise to many of the common practices we’ve seen in our combined 60 years in pediatrics.

In our effort to look at the evidence behind these beliefs, we encountered some limitations. For most of the recommendations we’ve put forth here, the best evidence that exists is category C—that is, recommendations based on consensus, usual practice, opinion, disease-oriented evidence, or case series for studies of diagnosis, treatment, prevention, or screening.

Though no randomized, controlled trials exist for most of the common practices reviewed here, we have included consensus statements from the best available sources, such as the Centers for Disease Control and Prevention and the AAP. Where appropriate, we also included clear, pathophysiologic reasoning to support each recommendation.

BELIEF 1: Breast milk is a complete nutritional source for a healthy term newborn

THE EVIDENCE: Breast milk is not a complete nutritional source for healthy term newborns. In fact, breast milk provides the ideal source of nutrition, and it is almost a complete and perfect source of nutrition—with one important exception. The AAP recommends that all breast-fed newborns receive 200 IU/day of vitamin D until they are getting at least 500 ml/day of Vitamin-D formula or milk (SOR: B).^1,2

The purpose of the supplementation is to prevent vitamin D deficiency and subsequent rickets. (Rickets does continue to occur in the United States.^3,4) The AAP makes no mention in its recommendation of infant pigmentation or the expected amount of exposure to sunshine. The AAP recommends that vitamin D supplementation begin by the time the infant is 2 months old.

BELIEF 2: Supplementing with formula because the mother’s milk hasn’t come in yet is a reasonable, routine practice

THE EVIDENCE: Formula supplements are not necessary as routine practice (SOR: C).

Formula supplements are counterproductive,⁵ taking away the primary stimulus for breast milk production—nursing at the breast. Infant dissatisfaction with the initial volume of breast milk produced actually works to the infant’s advantage,⁵ driving the child to the breast more often, and thus increasing the likelihood of successful breastfeeding.

In certain circumstances, formula supplements can be reasonable, such as when an infant is hypoglycemic or when the baby is receiving phototherapy⁵ and experiences excessive weight loss and becomes severely dehydrated. However, formula supplementation is not reasonable or necessary as routine practice.

BELIEF 3: Mothers on magnesium therapy should not breastfeed their infants

THE EVIDENCE: Mothers on magnesium therapy may continue to breastfeed their babies (SOR: B).

The misguided recommendation that mothers who are being treated with magnesium therapy should not breastfeed^6,7 is based on an unreasonable fear that magnesium therapy can cause hypermagnesemia in breastfed newborns due to excessive magnesium levels in the breast milk. Supplemental magnesium, usually given intravenously to mothers with severe preeclampsia, does not cross over into breast milk in any significant amount, even when the mother continues to need intravenous magnesium after the birth of her baby.^8,9

BELIEF 4: Mothers who are positive for hepatitis B surface antigen or who are carriers for hepatitis C should not breastfeed

THE EVIDENCE: Mothers who are hepatitis B surface antigen positive or carriers for hepatitis C can safely breastfeed their newborns¹⁰ (SOR: C).

The idea that mothers who are infected with hepatitis B or C should not breast-feed their babies at first seems obvious to many who care for newborns, as the diseases are transmitted through blood exposure, and nipple cracks with associated blood loss are common in mothers when they begin to breastfeed.

The hepatitis B immunization protocol for infants born to hepatitis B surface antigen positive mothers takes care of the first infectious concern.¹¹ In addition, no case of hepatitis C transmission from breast milk has ever been reported.¹¹ The Centers for Disease Control and Prevention confirms that the transmission rate of hepatitis C from infected mothers is the same whether the babies are breast- or bottle-fed.¹²

BELIEF 5: Mothers who are febrile should not breastfeed

THE EVIDENCE: In most cases, febrile mothers may safely breastfeed their infants (SOR: C).

The advice for mothers not to breast-feed while febrile seems intuitively true because of concern that the infection might pass over into the breast milk to the baby. This rarely happens. There are only 4 contraindications to breastfeeding during maternal fever:¹⁰

1. Active, untreated maternal tuberculosis
2. Mothers who are human T-cell lymphotropic virus type I or II positive
3. Mothers who are HIV-positive
4. Mothers with a herpes simplex lesion on the breast.

BELIEF 6: Mothers who smoke or drink alcohol should not breastfeed

THE EVIDENCE: While this recommendation seems self-evident, the research proving harmful effects to the infant is lacking¹³ (SOR: C).

In fact, in its most recent statement on “The Transfer of Drugs and Other Chemicals Into Human Milk,” the AAP removed nicotine from a table of drugs for which adverse effects have been reported in infants during breastfeeding.¹³

While it would be ideal if no breastfeeding mother smoked or drank alcohol, the fact of the matter is that some do. In light of this, it’s wise to encourage the mother to smoke outside the home, and to change her clothes before holding her baby. In so doing, she will avoid exposing her baby to most of the effects of secondhand smoke. In addition, while mothers who breastfeed their infants should, of course, avoid alcohol abuse, a single, occasional, small alcoholic drink is acceptable.¹⁴

BELIEF 7: Pacifiers are bad for newborns

THE EVIDENCE: It is not clear whether pacifiers are “bad” for newborns (SOR: C).

The belief that newborns should not have pacifiers came into being for a well-intended reason: Breastfeeding advocates were concerned that newborns would spend too much time sucking on the pacifier and too little time sucking at the breast, undermining the mother’s ability to breastfeed successfully.¹⁵

Consensus on the matter, though, is lacking. The UNICEF-World Health Organization Baby-Friendly Initiative, for instance, recommends that pacifiers not be used.¹⁶ The AAP, however, advises that pacifiers can be used once breastfeeding is well established.¹⁷

The research is also mixed. On the one hand, new evidence indicates that pacifier use may decrease the incidence of sudden infant death syndrome.¹⁸ On the other hand, pacifier use for longer than 48 months has been linked to orthodontic problems and dental caries.^19,20

Thus, while prolonged pacifier use may be harmful to dental hygiene, newer evidence allows that pacifiers may be acceptable in the first few years during breastfeeding.

BELIEF 8: Newborn emesis is an indication for a formula change

THE EVIDENCE: No literature supports the belief that it is appropriate to change an infant’s formula in response to emesis in the first 2 weeks of life (SOR: C).

The overwhelming majority of vomiting episodes in newborns have no accompanying medical problem.²¹ A 2002 study by Miyazawa et al that looked at more than 900 infants showed more than 47% of Japanese infants ≤1 month of age had daily occurrences of regurgitation or emesis without having an underlying medical disorder.²²

Newborns vomit for any number of benign reasons, including swallowed maternal blood or overfeeding.²¹ Gastroesophageal reflux can be a manifestation of milk allergy. However, a newborn infant is too young to manifest an antibody response to the protein in the formula. Therefore, switching formula because of vomiting due to milk allergy is not prudent in the first 10 to 14 days of life, the length of time needed to mount an antibody response to an antigen²³ and thus, the length of time needed to become allergic to an infant formula.

BELIEF 9: Umbilical cord care can prevent umbilical cord infections

THE EVIDENCE: There is no definitive evidence regarding the best method for preventing umbilical cord infections among babies living in developed countries (SOR: C).

In fact, there is no evidence that any topical preparation, be it a dye, an antiseptic, or an antibiotic, is any better at preventing umbilical cord infections than keeping the area clean and dry.²⁴ Umbilical cord infections such as omphalitis or tetanus neonatorum are more common in developing countries than high-income countries.¹⁶ In developed countries, cord care with topical antimicrobial agents is frequently unnecessary.²⁴

Infants who were delivered at home and those who room in with their mothers have no need of a topical antimicrobial therapy. If an infant is kept in a hospital nursery or intensive care unit, topical antimicrobial therapy to the cord may have some benefit in keeping down cord colonization with pathological bacteria such as methicillin-resistant Staphylococcus aureus.²⁴

Umbilical cord infections sometimes occur even when the cord area is kept clean and dry,²⁵ so healthcare providers must be attentive to signs of possible infection.¹⁶

BELIEF 10: It’s easy to spot when a newborn is jaundiced

THE EVIDENCE: Jaundice is actually difficult to detect in darkly pigmented babies,⁵ and in babies sent home within 24 hours of birth, because bilirubin levels reach maximum levels between the third and fourth days of life²⁶ (SOR: C).

Years ago, when infants stayed longer in the nursery, doctors had the chance to see them when their bilirubin level was highest and when the babies were most jaundiced. The current emphasis on early discharge does not allow this practice.

The AAP recommends clinical assessment of a newborn’s state of jaundice and that a bilirubin level be obtained whenever a physician is in doubt about the degree of clinical jaundice. The AAP also recommends that physicians consider obtaining a routine screening bilirubin in all newborns at the time of hospital discharge even if, by clinical assessment, the child is not jaundiced.⁵ (The AAP stopped short, though, of saying that such a screening test is required.) The AAP made these recommendations because of an increasing concern that the incidence of kernicterus in America is rising.²⁷

BELIEF 11: All infants who require phototherapy need IV fluids to prevent dehydration and enhance excretion of bilirubin

THE EVIDENCE: Unless the baby is clinically dehydrated, IV fluid therapy for infants under phototherapy is not needed²⁸ (SOR: c).

Though IV fluid therapy is commonly used to increase the excretion of bilirubin and combat dehydration, the research tells us that IV fluids do not bring down bilirubin levels and that even with mild dehydration, the best fluid therapy is breast milk or formula because it inhibits the enterohepatic circulation of bilirubin.⁵ Intravenous fluid therapy should be reserved for jaundiced newborns with moderate to severe dehydration, or those with mild dehydration⁵ who are not able to take fluid by mouth.

BELIEF 12: Breast-milk jaundice is best treated by stopping breastfeeding for 24 to 48 hours

THE EVIDENCE: breastfeeding should not be discontinued as a way to treat breast-milk jaundice (SOR: C).

In fact, breastfeeding should not be discontinued for jaundice due to any cause, as demonstrated in the opening scenario, unless you believe a baby is at risk of requiring an exchange transfusion. The need for phototherapy alone is not a sufficient reason to discontinue breastfeeding.⁵

Breast-milk jaundice is a common problem facing parents and physicians, but it is not a disease and does not represent an abnormality in and of itself. Rather, this normal physiologic condition gains its importance only in that it must be distinguished from pathological causes of newborn jaundice.²⁹ Breast-milk jaundice is believed to affect 1%²¹to 33%³⁰ of breastfed infants.

One treatment measure—to stop breastfeeding—began, in part, as a cost-effective way to diagnose breast-milk jaundice.³¹ Rechecking the bilirubin 24 to 48 hours after breastfeeding is discontinued would reveal a significant drop in the bilirubin level, confirming the diagnosis of breast-milk jaundice³² and obviating the need for testing for more serious medical illness.

The consequence of this misguided treatment approach, ie, discontinuing breastfeeding, is that some mothers are more likely to stop breastfeeding altogether.³³

Acknowledgments

The authors thank the following individuals for providing assistance with this manuscript: Mariateresa Esquivel, MD; Charlotte Hobbs, MD, PhD; Christopher Monnikendam, MD; and Clare Campbell Nesmith, MD.

Funding/Support

Dr Hall receives funding through DHS 1 P20 RR020146-01.

Correspondence
Bryan L. Burke, Jr., MD, FAAP, Associate Professor, Pediatrics and Neonatology, University of Arkansas for Medical Sciences, Arkansas Children’s Hospital, 4301 West Markham, Slot # 512-5B, Little Rock, AR 72205; [email protected]

Baby M is a 12-hour-old 40-week gestation boy, without any risk factors, whom you delivered vaginally to a first-time mother last evening. The following morning, his mother tells you that while she intends to breastfeed exclusively, she was told by a staff member in the nursery to give the baby some formula “until your milk comes in, so he won’t be so cranky.” The child appears entirely well.

The next morning, at 36-hours-of-age, Baby M still appears well, except that he is jaundiced. His bilirubin level is 15—just above the level at which phototherapy should be started, but well below the level at which an exchange transfusion is required. You order a transfer to the nursery for phototherapy.

You order IV fluids to prevent dehydration and to enhance the excretion of bilirubin. To bring his bilirubin level down more quickly, you advise the mother to temporarily stop breastfeeding and feed him a standard infant formula. The nursing staff gives the mother a breast pump to use until breastfeeding is re-started.

Forty-eight hours later Baby M is ready for discharge, and the mother has questions about infant feeding practices. You and the nurse explain to the mother that “breast milk provides complete nutrition” for her baby and that the baby will need no other nutrition until he starts solid foods at around 4 to 6 months of age.

Routine care, but was it appropriate?

The care you provided to Baby M and the advice you offered his mother might seem pretty routine, but look again. Was it appropriate?

In fact, your care—and that of the nursing staff—diverged from the evidence in 4 ways:

1. There was no need to give Baby M formula until his mother’s milk came in (Strength of Recommendation [SOR]: C).
2. It was not necessary to use IV fluids to treat jaundice in this otherwise healthy term newborn (SOR: C). IV fluids do not bring down bilirubin levels, and even with mild dehydration, the best fluid therapy is breast milk or formula.
3. There was no need to temporarily discontinue breastfeeding during phototherapy to bring Baby M’s bilirubin level down more quickly (SOR: C). While doing so may bring levels down a bit faster, it is not worth the risk that the mother may opt to discontinue breastfeeding entirely.
4. You should not have told Baby M’s mother that “breast milk provides complete nutrition.” In fact, the American Academy of Pediatrics (AAP) recommends that all breast-fed newborns receive supplemental vitamin D (SOR: B).

12 common beliefs that require a second look

Medicine has a long history of promoting practices like the ones described here, which are either not based upon scientific evidence or are in direct contradiction to existing evidence. Often, we do not realize how much of our advice is based on misunderstandings passed down from our teachers and absorbed from the cultural milieu in which we were raised.

In this article, we will review the most current evidence regarding 12 commonly held beliefs regarding newborn care. These beliefs relate to breast milk, breastfeeding, pacifier use, emesis, umbilical cord care, and jaundice. We’ve included them because these beliefs give rise to many of the common practices we’ve seen in our combined 60 years in pediatrics.

In our effort to look at the evidence behind these beliefs, we encountered some limitations. For most of the recommendations we’ve put forth here, the best evidence that exists is category C—that is, recommendations based on consensus, usual practice, opinion, disease-oriented evidence, or case series for studies of diagnosis, treatment, prevention, or screening.

Though no randomized, controlled trials exist for most of the common practices reviewed here, we have included consensus statements from the best available sources, such as the Centers for Disease Control and Prevention and the AAP. Where appropriate, we also included clear, pathophysiologic reasoning to support each recommendation.

BELIEF 1: Breast milk is a complete nutritional source for a healthy term newborn

THE EVIDENCE: Breast milk is not a complete nutritional source for healthy term newborns. In fact, breast milk provides the ideal source of nutrition, and it is almost a complete and perfect source of nutrition—with one important exception. The AAP recommends that all breast-fed newborns receive 200 IU/day of vitamin D until they are getting at least 500 ml/day of Vitamin-D formula or milk (SOR: B).^1,2

The purpose of the supplementation is to prevent vitamin D deficiency and subsequent rickets. (Rickets does continue to occur in the United States.^3,4) The AAP makes no mention in its recommendation of infant pigmentation or the expected amount of exposure to sunshine. The AAP recommends that vitamin D supplementation begin by the time the infant is 2 months old.

BELIEF 2: Supplementing with formula because the mother’s milk hasn’t come in yet is a reasonable, routine practice

THE EVIDENCE: Formula supplements are not necessary as routine practice (SOR: C).

Formula supplements are counterproductive,⁵ taking away the primary stimulus for breast milk production—nursing at the breast. Infant dissatisfaction with the initial volume of breast milk produced actually works to the infant’s advantage,⁵ driving the child to the breast more often, and thus increasing the likelihood of successful breastfeeding.

In certain circumstances, formula supplements can be reasonable, such as when an infant is hypoglycemic or when the baby is receiving phototherapy⁵ and experiences excessive weight loss and becomes severely dehydrated. However, formula supplementation is not reasonable or necessary as routine practice.

BELIEF 3: Mothers on magnesium therapy should not breastfeed their infants

THE EVIDENCE: Mothers on magnesium therapy may continue to breastfeed their babies (SOR: B).

The misguided recommendation that mothers who are being treated with magnesium therapy should not breastfeed^6,7 is based on an unreasonable fear that magnesium therapy can cause hypermagnesemia in breastfed newborns due to excessive magnesium levels in the breast milk. Supplemental magnesium, usually given intravenously to mothers with severe preeclampsia, does not cross over into breast milk in any significant amount, even when the mother continues to need intravenous magnesium after the birth of her baby.^8,9

BELIEF 4: Mothers who are positive for hepatitis B surface antigen or who are carriers for hepatitis C should not breastfeed

THE EVIDENCE: Mothers who are hepatitis B surface antigen positive or carriers for hepatitis C can safely breastfeed their newborns¹⁰ (SOR: C).

The idea that mothers who are infected with hepatitis B or C should not breast-feed their babies at first seems obvious to many who care for newborns, as the diseases are transmitted through blood exposure, and nipple cracks with associated blood loss are common in mothers when they begin to breastfeed.

The hepatitis B immunization protocol for infants born to hepatitis B surface antigen positive mothers takes care of the first infectious concern.¹¹ In addition, no case of hepatitis C transmission from breast milk has ever been reported.¹¹ The Centers for Disease Control and Prevention confirms that the transmission rate of hepatitis C from infected mothers is the same whether the babies are breast- or bottle-fed.¹²

BELIEF 5: Mothers who are febrile should not breastfeed

THE EVIDENCE: In most cases, febrile mothers may safely breastfeed their infants (SOR: C).

The advice for mothers not to breast-feed while febrile seems intuitively true because of concern that the infection might pass over into the breast milk to the baby. This rarely happens. There are only 4 contraindications to breastfeeding during maternal fever:¹⁰

1. Active, untreated maternal tuberculosis
2. Mothers who are human T-cell lymphotropic virus type I or II positive
3. Mothers who are HIV-positive
4. Mothers with a herpes simplex lesion on the breast.

BELIEF 6: Mothers who smoke or drink alcohol should not breastfeed

THE EVIDENCE: While this recommendation seems self-evident, the research proving harmful effects to the infant is lacking¹³ (SOR: C).

In fact, in its most recent statement on “The Transfer of Drugs and Other Chemicals Into Human Milk,” the AAP removed nicotine from a table of drugs for which adverse effects have been reported in infants during breastfeeding.¹³

While it would be ideal if no breastfeeding mother smoked or drank alcohol, the fact of the matter is that some do. In light of this, it’s wise to encourage the mother to smoke outside the home, and to change her clothes before holding her baby. In so doing, she will avoid exposing her baby to most of the effects of secondhand smoke. In addition, while mothers who breastfeed their infants should, of course, avoid alcohol abuse, a single, occasional, small alcoholic drink is acceptable.¹⁴

BELIEF 7: Pacifiers are bad for newborns

THE EVIDENCE: It is not clear whether pacifiers are “bad” for newborns (SOR: C).

The belief that newborns should not have pacifiers came into being for a well-intended reason: Breastfeeding advocates were concerned that newborns would spend too much time sucking on the pacifier and too little time sucking at the breast, undermining the mother’s ability to breastfeed successfully.¹⁵

Consensus on the matter, though, is lacking. The UNICEF-World Health Organization Baby-Friendly Initiative, for instance, recommends that pacifiers not be used.¹⁶ The AAP, however, advises that pacifiers can be used once breastfeeding is well established.¹⁷

The research is also mixed. On the one hand, new evidence indicates that pacifier use may decrease the incidence of sudden infant death syndrome.¹⁸ On the other hand, pacifier use for longer than 48 months has been linked to orthodontic problems and dental caries.^19,20

Thus, while prolonged pacifier use may be harmful to dental hygiene, newer evidence allows that pacifiers may be acceptable in the first few years during breastfeeding.

BELIEF 8: Newborn emesis is an indication for a formula change

THE EVIDENCE: No literature supports the belief that it is appropriate to change an infant’s formula in response to emesis in the first 2 weeks of life (SOR: C).

The overwhelming majority of vomiting episodes in newborns have no accompanying medical problem.²¹ A 2002 study by Miyazawa et al that looked at more than 900 infants showed more than 47% of Japanese infants ≤1 month of age had daily occurrences of regurgitation or emesis without having an underlying medical disorder.²²

Newborns vomit for any number of benign reasons, including swallowed maternal blood or overfeeding.²¹ Gastroesophageal reflux can be a manifestation of milk allergy. However, a newborn infant is too young to manifest an antibody response to the protein in the formula. Therefore, switching formula because of vomiting due to milk allergy is not prudent in the first 10 to 14 days of life, the length of time needed to mount an antibody response to an antigen²³ and thus, the length of time needed to become allergic to an infant formula.

BELIEF 9: Umbilical cord care can prevent umbilical cord infections

THE EVIDENCE: There is no definitive evidence regarding the best method for preventing umbilical cord infections among babies living in developed countries (SOR: C).

In fact, there is no evidence that any topical preparation, be it a dye, an antiseptic, or an antibiotic, is any better at preventing umbilical cord infections than keeping the area clean and dry.²⁴ Umbilical cord infections such as omphalitis or tetanus neonatorum are more common in developing countries than high-income countries.¹⁶ In developed countries, cord care with topical antimicrobial agents is frequently unnecessary.²⁴

Infants who were delivered at home and those who room in with their mothers have no need of a topical antimicrobial therapy. If an infant is kept in a hospital nursery or intensive care unit, topical antimicrobial therapy to the cord may have some benefit in keeping down cord colonization with pathological bacteria such as methicillin-resistant Staphylococcus aureus.²⁴

Umbilical cord infections sometimes occur even when the cord area is kept clean and dry,²⁵ so healthcare providers must be attentive to signs of possible infection.¹⁶

BELIEF 10: It’s easy to spot when a newborn is jaundiced

THE EVIDENCE: Jaundice is actually difficult to detect in darkly pigmented babies,⁵ and in babies sent home within 24 hours of birth, because bilirubin levels reach maximum levels between the third and fourth days of life²⁶ (SOR: C).

Years ago, when infants stayed longer in the nursery, doctors had the chance to see them when their bilirubin level was highest and when the babies were most jaundiced. The current emphasis on early discharge does not allow this practice.

The AAP recommends clinical assessment of a newborn’s state of jaundice and that a bilirubin level be obtained whenever a physician is in doubt about the degree of clinical jaundice. The AAP also recommends that physicians consider obtaining a routine screening bilirubin in all newborns at the time of hospital discharge even if, by clinical assessment, the child is not jaundiced.⁵ (The AAP stopped short, though, of saying that such a screening test is required.) The AAP made these recommendations because of an increasing concern that the incidence of kernicterus in America is rising.²⁷

BELIEF 11: All infants who require phototherapy need IV fluids to prevent dehydration and enhance excretion of bilirubin

THE EVIDENCE: Unless the baby is clinically dehydrated, IV fluid therapy for infants under phototherapy is not needed²⁸ (SOR: c).

Though IV fluid therapy is commonly used to increase the excretion of bilirubin and combat dehydration, the research tells us that IV fluids do not bring down bilirubin levels and that even with mild dehydration, the best fluid therapy is breast milk or formula because it inhibits the enterohepatic circulation of bilirubin.⁵ Intravenous fluid therapy should be reserved for jaundiced newborns with moderate to severe dehydration, or those with mild dehydration⁵ who are not able to take fluid by mouth.

BELIEF 12: Breast-milk jaundice is best treated by stopping breastfeeding for 24 to 48 hours

THE EVIDENCE: breastfeeding should not be discontinued as a way to treat breast-milk jaundice (SOR: C).

In fact, breastfeeding should not be discontinued for jaundice due to any cause, as demonstrated in the opening scenario, unless you believe a baby is at risk of requiring an exchange transfusion. The need for phototherapy alone is not a sufficient reason to discontinue breastfeeding.⁵

Breast-milk jaundice is a common problem facing parents and physicians, but it is not a disease and does not represent an abnormality in and of itself. Rather, this normal physiologic condition gains its importance only in that it must be distinguished from pathological causes of newborn jaundice.²⁹ Breast-milk jaundice is believed to affect 1%²¹to 33%³⁰ of breastfed infants.

One treatment measure—to stop breastfeeding—began, in part, as a cost-effective way to diagnose breast-milk jaundice.³¹ Rechecking the bilirubin 24 to 48 hours after breastfeeding is discontinued would reveal a significant drop in the bilirubin level, confirming the diagnosis of breast-milk jaundice³² and obviating the need for testing for more serious medical illness.

The consequence of this misguided treatment approach, ie, discontinuing breastfeeding, is that some mothers are more likely to stop breastfeeding altogether.³³

Acknowledgments

The authors thank the following individuals for providing assistance with this manuscript: Mariateresa Esquivel, MD; Charlotte Hobbs, MD, PhD; Christopher Monnikendam, MD; and Clare Campbell Nesmith, MD.

Funding/Support

Dr Hall receives funding through DHS 1 P20 RR020146-01.

Correspondence
Bryan L. Burke, Jr., MD, FAAP, Associate Professor, Pediatrics and Neonatology, University of Arkansas for Medical Sciences, Arkansas Children’s Hospital, 4301 West Markham, Slot # 512-5B, Little Rock, AR 72205; [email protected]

Baby M is a 12-hour-old 40-week gestation boy, without any risk factors, whom you delivered vaginally to a first-time mother last evening. The following morning, his mother tells you that while she intends to breastfeed exclusively, she was told by a staff member in the nursery to give the baby some formula “until your milk comes in, so he won’t be so cranky.” The child appears entirely well.

The next morning, at 36-hours-of-age, Baby M still appears well, except that he is jaundiced. His bilirubin level is 15—just above the level at which phototherapy should be started, but well below the level at which an exchange transfusion is required. You order a transfer to the nursery for phototherapy.

You order IV fluids to prevent dehydration and to enhance the excretion of bilirubin. To bring his bilirubin level down more quickly, you advise the mother to temporarily stop breastfeeding and feed him a standard infant formula. The nursing staff gives the mother a breast pump to use until breastfeeding is re-started.

Forty-eight hours later Baby M is ready for discharge, and the mother has questions about infant feeding practices. You and the nurse explain to the mother that “breast milk provides complete nutrition” for her baby and that the baby will need no other nutrition until he starts solid foods at around 4 to 6 months of age.

Routine care, but was it appropriate?

The care you provided to Baby M and the advice you offered his mother might seem pretty routine, but look again. Was it appropriate?

In fact, your care—and that of the nursing staff—diverged from the evidence in 4 ways:

1. There was no need to give Baby M formula until his mother’s milk came in (Strength of Recommendation [SOR]: C).
2. It was not necessary to use IV fluids to treat jaundice in this otherwise healthy term newborn (SOR: C). IV fluids do not bring down bilirubin levels, and even with mild dehydration, the best fluid therapy is breast milk or formula.
3. There was no need to temporarily discontinue breastfeeding during phototherapy to bring Baby M’s bilirubin level down more quickly (SOR: C). While doing so may bring levels down a bit faster, it is not worth the risk that the mother may opt to discontinue breastfeeding entirely.
4. You should not have told Baby M’s mother that “breast milk provides complete nutrition.” In fact, the American Academy of Pediatrics (AAP) recommends that all breast-fed newborns receive supplemental vitamin D (SOR: B).

12 common beliefs that require a second look

Medicine has a long history of promoting practices like the ones described here, which are either not based upon scientific evidence or are in direct contradiction to existing evidence. Often, we do not realize how much of our advice is based on misunderstandings passed down from our teachers and absorbed from the cultural milieu in which we were raised.

In this article, we will review the most current evidence regarding 12 commonly held beliefs regarding newborn care. These beliefs relate to breast milk, breastfeeding, pacifier use, emesis, umbilical cord care, and jaundice. We’ve included them because these beliefs give rise to many of the common practices we’ve seen in our combined 60 years in pediatrics.

In our effort to look at the evidence behind these beliefs, we encountered some limitations. For most of the recommendations we’ve put forth here, the best evidence that exists is category C—that is, recommendations based on consensus, usual practice, opinion, disease-oriented evidence, or case series for studies of diagnosis, treatment, prevention, or screening.

Though no randomized, controlled trials exist for most of the common practices reviewed here, we have included consensus statements from the best available sources, such as the Centers for Disease Control and Prevention and the AAP. Where appropriate, we also included clear, pathophysiologic reasoning to support each recommendation.

BELIEF 1: Breast milk is a complete nutritional source for a healthy term newborn

THE EVIDENCE: Breast milk is not a complete nutritional source for healthy term newborns. In fact, breast milk provides the ideal source of nutrition, and it is almost a complete and perfect source of nutrition—with one important exception. The AAP recommends that all breast-fed newborns receive 200 IU/day of vitamin D until they are getting at least 500 ml/day of Vitamin-D formula or milk (SOR: B).^1,2

The purpose of the supplementation is to prevent vitamin D deficiency and subsequent rickets. (Rickets does continue to occur in the United States.^3,4) The AAP makes no mention in its recommendation of infant pigmentation or the expected amount of exposure to sunshine. The AAP recommends that vitamin D supplementation begin by the time the infant is 2 months old.

BELIEF 2: Supplementing with formula because the mother’s milk hasn’t come in yet is a reasonable, routine practice

THE EVIDENCE: Formula supplements are not necessary as routine practice (SOR: C).

Formula supplements are counterproductive,⁵ taking away the primary stimulus for breast milk production—nursing at the breast. Infant dissatisfaction with the initial volume of breast milk produced actually works to the infant’s advantage,⁵ driving the child to the breast more often, and thus increasing the likelihood of successful breastfeeding.

In certain circumstances, formula supplements can be reasonable, such as when an infant is hypoglycemic or when the baby is receiving phototherapy⁵ and experiences excessive weight loss and becomes severely dehydrated. However, formula supplementation is not reasonable or necessary as routine practice.

BELIEF 3: Mothers on magnesium therapy should not breastfeed their infants

THE EVIDENCE: Mothers on magnesium therapy may continue to breastfeed their babies (SOR: B).

The misguided recommendation that mothers who are being treated with magnesium therapy should not breastfeed^6,7 is based on an unreasonable fear that magnesium therapy can cause hypermagnesemia in breastfed newborns due to excessive magnesium levels in the breast milk. Supplemental magnesium, usually given intravenously to mothers with severe preeclampsia, does not cross over into breast milk in any significant amount, even when the mother continues to need intravenous magnesium after the birth of her baby.^8,9

BELIEF 4: Mothers who are positive for hepatitis B surface antigen or who are carriers for hepatitis C should not breastfeed

THE EVIDENCE: Mothers who are hepatitis B surface antigen positive or carriers for hepatitis C can safely breastfeed their newborns¹⁰ (SOR: C).

The idea that mothers who are infected with hepatitis B or C should not breast-feed their babies at first seems obvious to many who care for newborns, as the diseases are transmitted through blood exposure, and nipple cracks with associated blood loss are common in mothers when they begin to breastfeed.

The hepatitis B immunization protocol for infants born to hepatitis B surface antigen positive mothers takes care of the first infectious concern.¹¹ In addition, no case of hepatitis C transmission from breast milk has ever been reported.¹¹ The Centers for Disease Control and Prevention confirms that the transmission rate of hepatitis C from infected mothers is the same whether the babies are breast- or bottle-fed.¹²

BELIEF 5: Mothers who are febrile should not breastfeed

THE EVIDENCE: In most cases, febrile mothers may safely breastfeed their infants (SOR: C).

The advice for mothers not to breast-feed while febrile seems intuitively true because of concern that the infection might pass over into the breast milk to the baby. This rarely happens. There are only 4 contraindications to breastfeeding during maternal fever:¹⁰

1. Active, untreated maternal tuberculosis
2. Mothers who are human T-cell lymphotropic virus type I or II positive
3. Mothers who are HIV-positive
4. Mothers with a herpes simplex lesion on the breast.

BELIEF 6: Mothers who smoke or drink alcohol should not breastfeed

THE EVIDENCE: While this recommendation seems self-evident, the research proving harmful effects to the infant is lacking¹³ (SOR: C).

In fact, in its most recent statement on “The Transfer of Drugs and Other Chemicals Into Human Milk,” the AAP removed nicotine from a table of drugs for which adverse effects have been reported in infants during breastfeeding.¹³

While it would be ideal if no breastfeeding mother smoked or drank alcohol, the fact of the matter is that some do. In light of this, it’s wise to encourage the mother to smoke outside the home, and to change her clothes before holding her baby. In so doing, she will avoid exposing her baby to most of the effects of secondhand smoke. In addition, while mothers who breastfeed their infants should, of course, avoid alcohol abuse, a single, occasional, small alcoholic drink is acceptable.¹⁴

BELIEF 7: Pacifiers are bad for newborns

THE EVIDENCE: It is not clear whether pacifiers are “bad” for newborns (SOR: C).

The belief that newborns should not have pacifiers came into being for a well-intended reason: Breastfeeding advocates were concerned that newborns would spend too much time sucking on the pacifier and too little time sucking at the breast, undermining the mother’s ability to breastfeed successfully.¹⁵

Consensus on the matter, though, is lacking. The UNICEF-World Health Organization Baby-Friendly Initiative, for instance, recommends that pacifiers not be used.¹⁶ The AAP, however, advises that pacifiers can be used once breastfeeding is well established.¹⁷

The research is also mixed. On the one hand, new evidence indicates that pacifier use may decrease the incidence of sudden infant death syndrome.¹⁸ On the other hand, pacifier use for longer than 48 months has been linked to orthodontic problems and dental caries.^19,20

Thus, while prolonged pacifier use may be harmful to dental hygiene, newer evidence allows that pacifiers may be acceptable in the first few years during breastfeeding.

BELIEF 8: Newborn emesis is an indication for a formula change

THE EVIDENCE: No literature supports the belief that it is appropriate to change an infant’s formula in response to emesis in the first 2 weeks of life (SOR: C).

The overwhelming majority of vomiting episodes in newborns have no accompanying medical problem.²¹ A 2002 study by Miyazawa et al that looked at more than 900 infants showed more than 47% of Japanese infants ≤1 month of age had daily occurrences of regurgitation or emesis without having an underlying medical disorder.²²

Newborns vomit for any number of benign reasons, including swallowed maternal blood or overfeeding.²¹ Gastroesophageal reflux can be a manifestation of milk allergy. However, a newborn infant is too young to manifest an antibody response to the protein in the formula. Therefore, switching formula because of vomiting due to milk allergy is not prudent in the first 10 to 14 days of life, the length of time needed to mount an antibody response to an antigen²³ and thus, the length of time needed to become allergic to an infant formula.

BELIEF 9: Umbilical cord care can prevent umbilical cord infections

THE EVIDENCE: There is no definitive evidence regarding the best method for preventing umbilical cord infections among babies living in developed countries (SOR: C).

In fact, there is no evidence that any topical preparation, be it a dye, an antiseptic, or an antibiotic, is any better at preventing umbilical cord infections than keeping the area clean and dry.²⁴ Umbilical cord infections such as omphalitis or tetanus neonatorum are more common in developing countries than high-income countries.¹⁶ In developed countries, cord care with topical antimicrobial agents is frequently unnecessary.²⁴

Infants who were delivered at home and those who room in with their mothers have no need of a topical antimicrobial therapy. If an infant is kept in a hospital nursery or intensive care unit, topical antimicrobial therapy to the cord may have some benefit in keeping down cord colonization with pathological bacteria such as methicillin-resistant Staphylococcus aureus.²⁴

Umbilical cord infections sometimes occur even when the cord area is kept clean and dry,²⁵ so healthcare providers must be attentive to signs of possible infection.¹⁶

BELIEF 10: It’s easy to spot when a newborn is jaundiced

THE EVIDENCE: Jaundice is actually difficult to detect in darkly pigmented babies,⁵ and in babies sent home within 24 hours of birth, because bilirubin levels reach maximum levels between the third and fourth days of life²⁶ (SOR: C).

Years ago, when infants stayed longer in the nursery, doctors had the chance to see them when their bilirubin level was highest and when the babies were most jaundiced. The current emphasis on early discharge does not allow this practice.

The AAP recommends clinical assessment of a newborn’s state of jaundice and that a bilirubin level be obtained whenever a physician is in doubt about the degree of clinical jaundice. The AAP also recommends that physicians consider obtaining a routine screening bilirubin in all newborns at the time of hospital discharge even if, by clinical assessment, the child is not jaundiced.⁵ (The AAP stopped short, though, of saying that such a screening test is required.) The AAP made these recommendations because of an increasing concern that the incidence of kernicterus in America is rising.²⁷

BELIEF 11: All infants who require phototherapy need IV fluids to prevent dehydration and enhance excretion of bilirubin

THE EVIDENCE: Unless the baby is clinically dehydrated, IV fluid therapy for infants under phototherapy is not needed²⁸ (SOR: c).

Though IV fluid therapy is commonly used to increase the excretion of bilirubin and combat dehydration, the research tells us that IV fluids do not bring down bilirubin levels and that even with mild dehydration, the best fluid therapy is breast milk or formula because it inhibits the enterohepatic circulation of bilirubin.⁵ Intravenous fluid therapy should be reserved for jaundiced newborns with moderate to severe dehydration, or those with mild dehydration⁵ who are not able to take fluid by mouth.

BELIEF 12: Breast-milk jaundice is best treated by stopping breastfeeding for 24 to 48 hours

THE EVIDENCE: breastfeeding should not be discontinued as a way to treat breast-milk jaundice (SOR: C).

In fact, breastfeeding should not be discontinued for jaundice due to any cause, as demonstrated in the opening scenario, unless you believe a baby is at risk of requiring an exchange transfusion. The need for phototherapy alone is not a sufficient reason to discontinue breastfeeding.⁵

Breast-milk jaundice is a common problem facing parents and physicians, but it is not a disease and does not represent an abnormality in and of itself. Rather, this normal physiologic condition gains its importance only in that it must be distinguished from pathological causes of newborn jaundice.²⁹ Breast-milk jaundice is believed to affect 1%²¹to 33%³⁰ of breastfed infants.

One treatment measure—to stop breastfeeding—began, in part, as a cost-effective way to diagnose breast-milk jaundice.³¹ Rechecking the bilirubin 24 to 48 hours after breastfeeding is discontinued would reveal a significant drop in the bilirubin level, confirming the diagnosis of breast-milk jaundice³² and obviating the need for testing for more serious medical illness.

The consequence of this misguided treatment approach, ie, discontinuing breastfeeding, is that some mothers are more likely to stop breastfeeding altogether.³³

Acknowledgments

The authors thank the following individuals for providing assistance with this manuscript: Mariateresa Esquivel, MD; Charlotte Hobbs, MD, PhD; Christopher Monnikendam, MD; and Clare Campbell Nesmith, MD.

Funding/Support

Dr Hall receives funding through DHS 1 P20 RR020146-01.

Correspondence
Bryan L. Burke, Jr., MD, FAAP, Associate Professor, Pediatrics and Neonatology, University of Arkansas for Medical Sciences, Arkansas Children’s Hospital, 4301 West Markham, Slot # 512-5B, Little Rock, AR 72205; [email protected]

More than 80% of patients with type 2 diabetes—including more than a third of patients with good metabolic control—have excessive postprandial hyperglycemia.¹ That’s unwelcome news for the 20 million Americans with type 2 diabetes, especially when you consider that post-prandial hyperglycemia is a strong independent risk factor for all-cause mortality and cardiovascular events.^2-5

To help our type 2 diabetes patients gain ideal control, we need to do at least 2 things better:

1. Measure and act on glycosylated hemoglobin (A_1c) levels.
2. Take a stepped approach to glycemic control, making full use of prandial insulin.

A_1clevels and the important role they play

Analysis of A_1c is the “gold standard” for monitoring glycemic control in patients with diabetes because it provides an indication of mean plasma glucose levels during the preceding 120 days.⁶ The relative contribution of fasting plasma glucose (FPG) and postprandial plasma glucose (PPG) to A_1c levels is a dynamic function of the extent of day-long hyperglycemia; FPG has a greater influence at higher A_1c levels and PPG has a predominant role at lower A_1c levels.⁷

The relationship between hyperglycemia, as measured by A_1c, and increased morbidity and mortality (including cardiovascular events) was demonstrated several years ago in the United Kingdom Prospective Diabetes Study (UKPDS).⁸ Interestingly, several studies have also found that fasting glucose levels alone are not a reliable predictor for hyperglycemia-related morbidity or mortality, whereas postprandial hyperglycemia, as noted in the introduction, is a strong independent risk factor for all-cause mortality and cardiovascular events.^2-5

Continued management of A_1c through tight control of both FPG and PPG may therefore improve patient long-term health outcomes. A_1c should be evaluated every 3 to 6 months, and appropriate changes to the patients’ treatment regimens should be made accordingly.

An algorithm for the stepwise approach

We typically use oral antidiabetic drugs typically are used as initial therapy for patients with newly diagnosed type 2 diabetes, especially those with initial A_1c levels of 6.0% to 8.0%.⁹ Three recent publications^10-12 provide an excellent analysis of the rationale for combination therapy to address multiple physiologic defects, as well as the relative efficacy of agents.

In 2006 the American Diabetes Association (ADA) and the European Association for the Study of Diabetes published a consensus statement that presented an algorithm for the initiation and adjustment of type 2 diabetes therapy (Figure).¹³ In this evidence-and experience-based treatment algorithm, the authors emphasize the achievement and maintenance of normal glycemic goals, initiating therapy with lifestyle intervention and metformin (Glucophage), not delaying therapy and transitioning to new regimens when glycemic targets are not achieved, and adding insulin therapy early to the regimens of patients who are not meeting glycemic targets.^9,13

You may also consider newer therapeutic options not included in the ADA’s 2007 treatment guidelines.¹⁴ Incretin mimetics and dipeptidyl-peptidase IV (DPP-IV) inhibitors are 2 new classes of antidiabetic agents that are effective for patients with type 2 diabetes. (See “2 new classes of antidiabetic agents: Incretin mimetics and DPP-IV inhibitors” ^15-22 ) Most patients with type 2 diabetes will, however, eventually require insulin therapy to maintain optimal glycemic control.²³

Advancement to insulin therapy

Basal insulin replacement achieves glycemic control

The addition of once-daily basal insulin to oral antidiabetic drug regimens is a simple way to introduce insulin therapy and achieve glycemic control. (See “A guide to basal insulin dosing and titration”^14,24-28 )

• In a randomized, parallel, multi-center study, treatment with once-daily insulin glargine (Lantus) or neutral protamine Hagedorn (NPH) insulin (Humulin, Novolin) added to preexisting oral antidiabetic drug regimens for 756 patients with inadequately controlled type 2 diabetes (A_1c >7.5%) effectively achieved the goal A_1c of ≤7.0% for most patients. More patients in the insulin glargine group were able to reach goal without experiencing any nocturnal hypoglycemia, compared with those in the NPH group (33.2% vs 26.7%, P<.05).²⁴
• A similar study used a forced-titration algorithm to compare NPH insulin with insulin detemir (Levemir).²⁶ This study demonstrated that a similar reduction in A_1c could be achieved with insulin detemir and NPH insulin over 24 weeks (1.8% and 1.9%, P=ns), but weight gain and nocturnal hypoglycemia incidence were significantly lower with insulin detemir compared with NPH insulin (1.2 kg vs 2.8 kg, and 160 vs 349 events, respectively, P<.001 for both).

Prandial insulin is as effective as carbohydrate counting

If a patient doesn’t reach his or her A_1c targets despite appropriate titration of the basal insulin dose, injections of a rapid-acting insulin analog at mealtime may be necessary. You can add rapid-acting insulin to the regimen, starting with 1 injection at the largest meal of the day and then adding an injection at additional meals as needed.

• A typical starting dose of rapid-acting prandial insulin (insulin aspart [NovoLog], insulin glulisine [Apidra], or insulin lispro [Humalog]) would be 5 to 10 U per meal.²⁹
• The range of daily dose, when used at 3 meals per day, would be 0.2 to 0.5 U/kg per day (ie, 0.1 to 0.15 U/kg per meal).²⁹ For a patient who weighs 100 kg, that would mean 20 U (6–7 units before each meal) per day.

FIGURE
An evidence-based algorithm for achieving normal glycemic goals in patients with type 2 diabetes

Recent data indicate that a simple treatment algorithm based on preprandial glucose patterns can be as effective as carbohydrate counting for the dose titration of prandial insulin.³⁰ In this 24-week study, insulin glulisine was added to basal-prandial insulin therapy, with insulin glargine as the basal insulin component. Glulisine was adjusted to target using either a simple algorithm of adding 1, 2, or 3 U based on premeal glucose patterns or standard carbohydrate counting. The carbohydrate counting–based dose adjustment and the algorithm-based titration treatment arms achieved similar A_1c reductions. However, patients using the simple algorithm experienced significantly less symptomatic hypoglycemia (P=.02).

Rapid-acting analogs allow more flexible administration

Prior to the development of rapid-acting insulin analogs, regular human insulin (RHI) was the only available insulin suitable for prandial glycemic control. However, it had significant limitations, including the need for it to be injected 30 to 45 minutes before eating (and the poor compliance with this requirement), variability in peak levels (between patients and with the same patient), variability in absorption based on injection site, and frequent episodes of hypoglycemia.^31,32

Newer rapid-acting insulin analogs such as insulin aspart, insulin glulisine, and insulin lispro demonstrate improved pharmacokinetic profiles with more rapid onset, faster time to peak activity, and shorter duration of action than RHI.^32,33 These rapid-acting analogs allow administration right before or right after a meal, resulting in improved glycemic control without increased hypoglycemia or weight gain.^34,35 Whereas the rapid onset of action of these analogs allows for administration 5 to 15 minutes before a meal, the patient can administer insulin glulisine within 20 minutes of the start of the meal.³⁶ The addition of just 1 dose of prandial insulin to existing basal insulin plus oral antidiabetic drug therapy offers patients a substantial benefit.³⁷

A new option: inhaled insulin

The US Food and Drug Administration recently approved an inhaled prandial insulin. Research has shown that it effectively addresses postprandial glucose excursions for patients with type 2 diabetes.^38,39 A 12-week trial comparing A_1c levels among patients switched to inhaled insulin (Exubera) before meals (n=76) or rosiglitazone (Avandia) 4 mg twice daily (n=69) found that inhaled insulin reduced A1c to a greater degree than rosiglitazone (–2.3% vs –1.4%); however, patients receiving inhaled insulin experienced a greater incidence of hypoglycemia (0.7 vs 0.05 episodes per subject-month).³⁹

Inhaled insulin can be used as monotherapy or in conjunction with oral agents or a long-acting basal insulin. Inhaled insulin has a rapid onset of action (within 10–20 minutes, comparable with rapid-acting insulin analogs) and a duration of glucose-lowering activity of approximately 6 hours (comparable with RHI).⁴⁰ This is useful for patients reluctant to begin insulin therapy because of injections; however, you will need to closely monitor hypoglycemia.

TABLE
How to use sensitivity factors to calculate 24-hour insulin need

Basal-prandial insulin in new type 2 diabetes

In certain cases, it may be more appropriate to initiate insulin therapy using a basal-prandial regimen that includes injections of prandial insulin with each meal of the day. Such cases include patients with newly diagnosed type 2 diabetes who have A_1c levels >10.0%, or insulin-naive patients on oral antidiabetic drug regimens who have A_1c levels >8.5%.²⁵

You can calculate the starting total 24-hour insulin dosage for both the basal and prandial insulin components by multiplying body weight in kg by a factor based on the patient’s estimated insulin sensitivity ( TABLE ).²⁹

Once you have this 24-hour insulin dose, you’ll then need to calculate the dose of basal insulin, which is 50% of the 24-hour total insulin dose, administered once daily. The remaining 50% of the total 24-hour dose provides prandial insulin coverage and is usually administered as follows:

• 30% to 40% at breakfast
• 30% at lunch
• 30% to 40% at dinner

Patients will need to adjust prandial insulin doses based on self-monitored blood glucose values.

Premixed insulin formulations

You should have your patients administer basal-prandial insulin as separate injections (eg, insulin glargine and insulin glulisine,³⁰ or insulin detemir and insulin aspart²⁵ ). The premixed (NPH based) formulations provide fixed doses of an intermediate-or long-acting insulin combined with a short-acting insulin. Although this method may be convenient to administer, it is more rigid and may not account for mealtimes and exercise. As a result, insulin levels will not match physiological insulin and thus, the risk for hypoglycemia increases. Another disadvantage is that adjustments to the dose based on self-monitored glucose levels are not possible with pre-mixed formulations.⁴¹

Separating the basal and prandial insulin components allows the insulin regimen to be adapted to an individual’s needs, thereby providing glycemic control with less propensity for hypoglycemia.

Acknowledgments

This article was supported by Sanofi-Aventis US. While the author is responsible for all content, he gratefully acknowledges the embryon scientific staff, who assisted in the preparation of a first draft of this article based on an author-approved outline, and also assisted in implementing author revisions.

Correspondence
George E. Dailey, MD, Senior Consultant, Division of Diabetes and Endocrinology, Head, Diabetes Research, Scripps Clinic, 10666 N. Torrey Pines Road, La Jolla, CA 92037; [email protected]

More than 80% of patients with type 2 diabetes—including more than a third of patients with good metabolic control—have excessive postprandial hyperglycemia.¹ That’s unwelcome news for the 20 million Americans with type 2 diabetes, especially when you consider that post-prandial hyperglycemia is a strong independent risk factor for all-cause mortality and cardiovascular events.^2-5

To help our type 2 diabetes patients gain ideal control, we need to do at least 2 things better:

1. Measure and act on glycosylated hemoglobin (A_1c) levels.
2. Take a stepped approach to glycemic control, making full use of prandial insulin.

A_1clevels and the important role they play

Analysis of A_1c is the “gold standard” for monitoring glycemic control in patients with diabetes because it provides an indication of mean plasma glucose levels during the preceding 120 days.⁶ The relative contribution of fasting plasma glucose (FPG) and postprandial plasma glucose (PPG) to A_1c levels is a dynamic function of the extent of day-long hyperglycemia; FPG has a greater influence at higher A_1c levels and PPG has a predominant role at lower A_1c levels.⁷

The relationship between hyperglycemia, as measured by A_1c, and increased morbidity and mortality (including cardiovascular events) was demonstrated several years ago in the United Kingdom Prospective Diabetes Study (UKPDS).⁸ Interestingly, several studies have also found that fasting glucose levels alone are not a reliable predictor for hyperglycemia-related morbidity or mortality, whereas postprandial hyperglycemia, as noted in the introduction, is a strong independent risk factor for all-cause mortality and cardiovascular events.^2-5

Continued management of A_1c through tight control of both FPG and PPG may therefore improve patient long-term health outcomes. A_1c should be evaluated every 3 to 6 months, and appropriate changes to the patients’ treatment regimens should be made accordingly.

An algorithm for the stepwise approach

We typically use oral antidiabetic drugs typically are used as initial therapy for patients with newly diagnosed type 2 diabetes, especially those with initial A_1c levels of 6.0% to 8.0%.⁹ Three recent publications^10-12 provide an excellent analysis of the rationale for combination therapy to address multiple physiologic defects, as well as the relative efficacy of agents.

In 2006 the American Diabetes Association (ADA) and the European Association for the Study of Diabetes published a consensus statement that presented an algorithm for the initiation and adjustment of type 2 diabetes therapy (Figure).¹³ In this evidence-and experience-based treatment algorithm, the authors emphasize the achievement and maintenance of normal glycemic goals, initiating therapy with lifestyle intervention and metformin (Glucophage), not delaying therapy and transitioning to new regimens when glycemic targets are not achieved, and adding insulin therapy early to the regimens of patients who are not meeting glycemic targets.^9,13

You may also consider newer therapeutic options not included in the ADA’s 2007 treatment guidelines.¹⁴ Incretin mimetics and dipeptidyl-peptidase IV (DPP-IV) inhibitors are 2 new classes of antidiabetic agents that are effective for patients with type 2 diabetes. (See “2 new classes of antidiabetic agents: Incretin mimetics and DPP-IV inhibitors” ^15-22 ) Most patients with type 2 diabetes will, however, eventually require insulin therapy to maintain optimal glycemic control.²³

Advancement to insulin therapy

Basal insulin replacement achieves glycemic control

The addition of once-daily basal insulin to oral antidiabetic drug regimens is a simple way to introduce insulin therapy and achieve glycemic control. (See “A guide to basal insulin dosing and titration”^14,24-28 )

• In a randomized, parallel, multi-center study, treatment with once-daily insulin glargine (Lantus) or neutral protamine Hagedorn (NPH) insulin (Humulin, Novolin) added to preexisting oral antidiabetic drug regimens for 756 patients with inadequately controlled type 2 diabetes (A_1c >7.5%) effectively achieved the goal A_1c of ≤7.0% for most patients. More patients in the insulin glargine group were able to reach goal without experiencing any nocturnal hypoglycemia, compared with those in the NPH group (33.2% vs 26.7%, P<.05).²⁴
• A similar study used a forced-titration algorithm to compare NPH insulin with insulin detemir (Levemir).²⁶ This study demonstrated that a similar reduction in A_1c could be achieved with insulin detemir and NPH insulin over 24 weeks (1.8% and 1.9%, P=ns), but weight gain and nocturnal hypoglycemia incidence were significantly lower with insulin detemir compared with NPH insulin (1.2 kg vs 2.8 kg, and 160 vs 349 events, respectively, P<.001 for both).

Prandial insulin is as effective as carbohydrate counting

If a patient doesn’t reach his or her A_1c targets despite appropriate titration of the basal insulin dose, injections of a rapid-acting insulin analog at mealtime may be necessary. You can add rapid-acting insulin to the regimen, starting with 1 injection at the largest meal of the day and then adding an injection at additional meals as needed.

• A typical starting dose of rapid-acting prandial insulin (insulin aspart [NovoLog], insulin glulisine [Apidra], or insulin lispro [Humalog]) would be 5 to 10 U per meal.²⁹
• The range of daily dose, when used at 3 meals per day, would be 0.2 to 0.5 U/kg per day (ie, 0.1 to 0.15 U/kg per meal).²⁹ For a patient who weighs 100 kg, that would mean 20 U (6–7 units before each meal) per day.

FIGURE
An evidence-based algorithm for achieving normal glycemic goals in patients with type 2 diabetes

Recent data indicate that a simple treatment algorithm based on preprandial glucose patterns can be as effective as carbohydrate counting for the dose titration of prandial insulin.³⁰ In this 24-week study, insulin glulisine was added to basal-prandial insulin therapy, with insulin glargine as the basal insulin component. Glulisine was adjusted to target using either a simple algorithm of adding 1, 2, or 3 U based on premeal glucose patterns or standard carbohydrate counting. The carbohydrate counting–based dose adjustment and the algorithm-based titration treatment arms achieved similar A_1c reductions. However, patients using the simple algorithm experienced significantly less symptomatic hypoglycemia (P=.02).

Rapid-acting analogs allow more flexible administration

Prior to the development of rapid-acting insulin analogs, regular human insulin (RHI) was the only available insulin suitable for prandial glycemic control. However, it had significant limitations, including the need for it to be injected 30 to 45 minutes before eating (and the poor compliance with this requirement), variability in peak levels (between patients and with the same patient), variability in absorption based on injection site, and frequent episodes of hypoglycemia.^31,32

Newer rapid-acting insulin analogs such as insulin aspart, insulin glulisine, and insulin lispro demonstrate improved pharmacokinetic profiles with more rapid onset, faster time to peak activity, and shorter duration of action than RHI.^32,33 These rapid-acting analogs allow administration right before or right after a meal, resulting in improved glycemic control without increased hypoglycemia or weight gain.^34,35 Whereas the rapid onset of action of these analogs allows for administration 5 to 15 minutes before a meal, the patient can administer insulin glulisine within 20 minutes of the start of the meal.³⁶ The addition of just 1 dose of prandial insulin to existing basal insulin plus oral antidiabetic drug therapy offers patients a substantial benefit.³⁷

A new option: inhaled insulin

The US Food and Drug Administration recently approved an inhaled prandial insulin. Research has shown that it effectively addresses postprandial glucose excursions for patients with type 2 diabetes.^38,39 A 12-week trial comparing A_1c levels among patients switched to inhaled insulin (Exubera) before meals (n=76) or rosiglitazone (Avandia) 4 mg twice daily (n=69) found that inhaled insulin reduced A1c to a greater degree than rosiglitazone (–2.3% vs –1.4%); however, patients receiving inhaled insulin experienced a greater incidence of hypoglycemia (0.7 vs 0.05 episodes per subject-month).³⁹

Inhaled insulin can be used as monotherapy or in conjunction with oral agents or a long-acting basal insulin. Inhaled insulin has a rapid onset of action (within 10–20 minutes, comparable with rapid-acting insulin analogs) and a duration of glucose-lowering activity of approximately 6 hours (comparable with RHI).⁴⁰ This is useful for patients reluctant to begin insulin therapy because of injections; however, you will need to closely monitor hypoglycemia.

TABLE
How to use sensitivity factors to calculate 24-hour insulin need

Basal-prandial insulin in new type 2 diabetes

In certain cases, it may be more appropriate to initiate insulin therapy using a basal-prandial regimen that includes injections of prandial insulin with each meal of the day. Such cases include patients with newly diagnosed type 2 diabetes who have A_1c levels >10.0%, or insulin-naive patients on oral antidiabetic drug regimens who have A_1c levels >8.5%.²⁵

You can calculate the starting total 24-hour insulin dosage for both the basal and prandial insulin components by multiplying body weight in kg by a factor based on the patient’s estimated insulin sensitivity ( TABLE ).²⁹

Once you have this 24-hour insulin dose, you’ll then need to calculate the dose of basal insulin, which is 50% of the 24-hour total insulin dose, administered once daily. The remaining 50% of the total 24-hour dose provides prandial insulin coverage and is usually administered as follows:

• 30% to 40% at breakfast
• 30% at lunch
• 30% to 40% at dinner

Patients will need to adjust prandial insulin doses based on self-monitored blood glucose values.

Premixed insulin formulations

You should have your patients administer basal-prandial insulin as separate injections (eg, insulin glargine and insulin glulisine,³⁰ or insulin detemir and insulin aspart²⁵ ). The premixed (NPH based) formulations provide fixed doses of an intermediate-or long-acting insulin combined with a short-acting insulin. Although this method may be convenient to administer, it is more rigid and may not account for mealtimes and exercise. As a result, insulin levels will not match physiological insulin and thus, the risk for hypoglycemia increases. Another disadvantage is that adjustments to the dose based on self-monitored glucose levels are not possible with pre-mixed formulations.⁴¹

Separating the basal and prandial insulin components allows the insulin regimen to be adapted to an individual’s needs, thereby providing glycemic control with less propensity for hypoglycemia.

Acknowledgments

This article was supported by Sanofi-Aventis US. While the author is responsible for all content, he gratefully acknowledges the embryon scientific staff, who assisted in the preparation of a first draft of this article based on an author-approved outline, and also assisted in implementing author revisions.

Correspondence
George E. Dailey, MD, Senior Consultant, Division of Diabetes and Endocrinology, Head, Diabetes Research, Scripps Clinic, 10666 N. Torrey Pines Road, La Jolla, CA 92037; [email protected]

More than 80% of patients with type 2 diabetes—including more than a third of patients with good metabolic control—have excessive postprandial hyperglycemia.¹ That’s unwelcome news for the 20 million Americans with type 2 diabetes, especially when you consider that post-prandial hyperglycemia is a strong independent risk factor for all-cause mortality and cardiovascular events.^2-5

To help our type 2 diabetes patients gain ideal control, we need to do at least 2 things better:

1. Measure and act on glycosylated hemoglobin (A_1c) levels.
2. Take a stepped approach to glycemic control, making full use of prandial insulin.

A_1clevels and the important role they play

Analysis of A_1c is the “gold standard” for monitoring glycemic control in patients with diabetes because it provides an indication of mean plasma glucose levels during the preceding 120 days.⁶ The relative contribution of fasting plasma glucose (FPG) and postprandial plasma glucose (PPG) to A_1c levels is a dynamic function of the extent of day-long hyperglycemia; FPG has a greater influence at higher A_1c levels and PPG has a predominant role at lower A_1c levels.⁷

The relationship between hyperglycemia, as measured by A_1c, and increased morbidity and mortality (including cardiovascular events) was demonstrated several years ago in the United Kingdom Prospective Diabetes Study (UKPDS).⁸ Interestingly, several studies have also found that fasting glucose levels alone are not a reliable predictor for hyperglycemia-related morbidity or mortality, whereas postprandial hyperglycemia, as noted in the introduction, is a strong independent risk factor for all-cause mortality and cardiovascular events.^2-5

Continued management of A_1c through tight control of both FPG and PPG may therefore improve patient long-term health outcomes. A_1c should be evaluated every 3 to 6 months, and appropriate changes to the patients’ treatment regimens should be made accordingly.

An algorithm for the stepwise approach

We typically use oral antidiabetic drugs typically are used as initial therapy for patients with newly diagnosed type 2 diabetes, especially those with initial A_1c levels of 6.0% to 8.0%.⁹ Three recent publications^10-12 provide an excellent analysis of the rationale for combination therapy to address multiple physiologic defects, as well as the relative efficacy of agents.

In 2006 the American Diabetes Association (ADA) and the European Association for the Study of Diabetes published a consensus statement that presented an algorithm for the initiation and adjustment of type 2 diabetes therapy (Figure).¹³ In this evidence-and experience-based treatment algorithm, the authors emphasize the achievement and maintenance of normal glycemic goals, initiating therapy with lifestyle intervention and metformin (Glucophage), not delaying therapy and transitioning to new regimens when glycemic targets are not achieved, and adding insulin therapy early to the regimens of patients who are not meeting glycemic targets.^9,13

You may also consider newer therapeutic options not included in the ADA’s 2007 treatment guidelines.¹⁴ Incretin mimetics and dipeptidyl-peptidase IV (DPP-IV) inhibitors are 2 new classes of antidiabetic agents that are effective for patients with type 2 diabetes. (See “2 new classes of antidiabetic agents: Incretin mimetics and DPP-IV inhibitors” ^15-22 ) Most patients with type 2 diabetes will, however, eventually require insulin therapy to maintain optimal glycemic control.²³

Advancement to insulin therapy

Basal insulin replacement achieves glycemic control

The addition of once-daily basal insulin to oral antidiabetic drug regimens is a simple way to introduce insulin therapy and achieve glycemic control. (See “A guide to basal insulin dosing and titration”^14,24-28 )

• In a randomized, parallel, multi-center study, treatment with once-daily insulin glargine (Lantus) or neutral protamine Hagedorn (NPH) insulin (Humulin, Novolin) added to preexisting oral antidiabetic drug regimens for 756 patients with inadequately controlled type 2 diabetes (A_1c >7.5%) effectively achieved the goal A_1c of ≤7.0% for most patients. More patients in the insulin glargine group were able to reach goal without experiencing any nocturnal hypoglycemia, compared with those in the NPH group (33.2% vs 26.7%, P<.05).²⁴
• A similar study used a forced-titration algorithm to compare NPH insulin with insulin detemir (Levemir).²⁶ This study demonstrated that a similar reduction in A_1c could be achieved with insulin detemir and NPH insulin over 24 weeks (1.8% and 1.9%, P=ns), but weight gain and nocturnal hypoglycemia incidence were significantly lower with insulin detemir compared with NPH insulin (1.2 kg vs 2.8 kg, and 160 vs 349 events, respectively, P<.001 for both).

Prandial insulin is as effective as carbohydrate counting

If a patient doesn’t reach his or her A_1c targets despite appropriate titration of the basal insulin dose, injections of a rapid-acting insulin analog at mealtime may be necessary. You can add rapid-acting insulin to the regimen, starting with 1 injection at the largest meal of the day and then adding an injection at additional meals as needed.

• A typical starting dose of rapid-acting prandial insulin (insulin aspart [NovoLog], insulin glulisine [Apidra], or insulin lispro [Humalog]) would be 5 to 10 U per meal.²⁹
• The range of daily dose, when used at 3 meals per day, would be 0.2 to 0.5 U/kg per day (ie, 0.1 to 0.15 U/kg per meal).²⁹ For a patient who weighs 100 kg, that would mean 20 U (6–7 units before each meal) per day.

FIGURE
An evidence-based algorithm for achieving normal glycemic goals in patients with type 2 diabetes

Recent data indicate that a simple treatment algorithm based on preprandial glucose patterns can be as effective as carbohydrate counting for the dose titration of prandial insulin.³⁰ In this 24-week study, insulin glulisine was added to basal-prandial insulin therapy, with insulin glargine as the basal insulin component. Glulisine was adjusted to target using either a simple algorithm of adding 1, 2, or 3 U based on premeal glucose patterns or standard carbohydrate counting. The carbohydrate counting–based dose adjustment and the algorithm-based titration treatment arms achieved similar A_1c reductions. However, patients using the simple algorithm experienced significantly less symptomatic hypoglycemia (P=.02).

Rapid-acting analogs allow more flexible administration

Prior to the development of rapid-acting insulin analogs, regular human insulin (RHI) was the only available insulin suitable for prandial glycemic control. However, it had significant limitations, including the need for it to be injected 30 to 45 minutes before eating (and the poor compliance with this requirement), variability in peak levels (between patients and with the same patient), variability in absorption based on injection site, and frequent episodes of hypoglycemia.^31,32

Newer rapid-acting insulin analogs such as insulin aspart, insulin glulisine, and insulin lispro demonstrate improved pharmacokinetic profiles with more rapid onset, faster time to peak activity, and shorter duration of action than RHI.^32,33 These rapid-acting analogs allow administration right before or right after a meal, resulting in improved glycemic control without increased hypoglycemia or weight gain.^34,35 Whereas the rapid onset of action of these analogs allows for administration 5 to 15 minutes before a meal, the patient can administer insulin glulisine within 20 minutes of the start of the meal.³⁶ The addition of just 1 dose of prandial insulin to existing basal insulin plus oral antidiabetic drug therapy offers patients a substantial benefit.³⁷

A new option: inhaled insulin

The US Food and Drug Administration recently approved an inhaled prandial insulin. Research has shown that it effectively addresses postprandial glucose excursions for patients with type 2 diabetes.^38,39 A 12-week trial comparing A_1c levels among patients switched to inhaled insulin (Exubera) before meals (n=76) or rosiglitazone (Avandia) 4 mg twice daily (n=69) found that inhaled insulin reduced A1c to a greater degree than rosiglitazone (–2.3% vs –1.4%); however, patients receiving inhaled insulin experienced a greater incidence of hypoglycemia (0.7 vs 0.05 episodes per subject-month).³⁹

Inhaled insulin can be used as monotherapy or in conjunction with oral agents or a long-acting basal insulin. Inhaled insulin has a rapid onset of action (within 10–20 minutes, comparable with rapid-acting insulin analogs) and a duration of glucose-lowering activity of approximately 6 hours (comparable with RHI).⁴⁰ This is useful for patients reluctant to begin insulin therapy because of injections; however, you will need to closely monitor hypoglycemia.

TABLE
How to use sensitivity factors to calculate 24-hour insulin need

Basal-prandial insulin in new type 2 diabetes

In certain cases, it may be more appropriate to initiate insulin therapy using a basal-prandial regimen that includes injections of prandial insulin with each meal of the day. Such cases include patients with newly diagnosed type 2 diabetes who have A_1c levels >10.0%, or insulin-naive patients on oral antidiabetic drug regimens who have A_1c levels >8.5%.²⁵

You can calculate the starting total 24-hour insulin dosage for both the basal and prandial insulin components by multiplying body weight in kg by a factor based on the patient’s estimated insulin sensitivity ( TABLE ).²⁹

Once you have this 24-hour insulin dose, you’ll then need to calculate the dose of basal insulin, which is 50% of the 24-hour total insulin dose, administered once daily. The remaining 50% of the total 24-hour dose provides prandial insulin coverage and is usually administered as follows:

• 30% to 40% at breakfast
• 30% at lunch
• 30% to 40% at dinner

Patients will need to adjust prandial insulin doses based on self-monitored blood glucose values.

Premixed insulin formulations

You should have your patients administer basal-prandial insulin as separate injections (eg, insulin glargine and insulin glulisine,³⁰ or insulin detemir and insulin aspart²⁵ ). The premixed (NPH based) formulations provide fixed doses of an intermediate-or long-acting insulin combined with a short-acting insulin. Although this method may be convenient to administer, it is more rigid and may not account for mealtimes and exercise. As a result, insulin levels will not match physiological insulin and thus, the risk for hypoglycemia increases. Another disadvantage is that adjustments to the dose based on self-monitored glucose levels are not possible with pre-mixed formulations.⁴¹

Separating the basal and prandial insulin components allows the insulin regimen to be adapted to an individual’s needs, thereby providing glycemic control with less propensity for hypoglycemia.

Acknowledgments

This article was supported by Sanofi-Aventis US. While the author is responsible for all content, he gratefully acknowledges the embryon scientific staff, who assisted in the preparation of a first draft of this article based on an author-approved outline, and also assisted in implementing author revisions.

Correspondence
George E. Dailey, MD, Senior Consultant, Division of Diabetes and Endocrinology, Head, Diabetes Research, Scripps Clinic, 10666 N. Torrey Pines Road, La Jolla, CA 92037; [email protected]

The 2007 guidelines from the Infectious Diseases Society of America (IDSA)/American Thoracic Society (ATS)¹ are a blend of level-of-evidence strength and consensus opinion—a unified, evidence-based document. these new recommendations address prior discrepancies between the 2 specialties. We developed a CAP treatment algorithm based on the new advisory. (The following text includes levels of evidence.)

Site-of-care decisions

1. Let severity score be your guide

Based on evidence that physicians often hospitalize patients for CAP who could be managed as outpatients, the new guidelines recommend that we use an illness severity score (strong recommendation, level I evidence).

Previous guidelines advised only that we consider using a severity score.

• Use the validated Pneumonia Severity Index or the easier-to-use CURB-65. Patients with a CURB-65 score of 2 or more generally require hospitalization (moderate recommendation, level III evidence).
• Ability to reliably and safely take medications at home must also be taken into account (strong recommendation, level II evidence.)

2. Admit to ICU promptly if needed

The criteria for admission to the ICU is similar to the previous ATS guidelines, but the list of minor criteria is more extensive. This change reflects evidence demonstrating worse outcomes in patients whose transfer to the ICU was delayed. This new criteria has not been validated.

• Patients requiring vasopressors for blood pressure support or with hypoxemic respiratory failure should be admitted to the ICU—these are major criteria (strong recommendation, level II evidence).
• Patients with 3 or more minor criteria should also be directly admitted to the ICU (moderate recommendation, level II evidence).

3. Identify who needs more tests

In the wake of controversy about diagnostic testing recommendations, the new guidelines attempt to better identify patients who would benefit from further testing (TABLE).

• 12 indications. Prior ATS guidelines lacked specifics on required additional testing, but the new guidelines give 12 clinical indications for more extensive evaluation, and identify which tests are recommended for each indication (strong recommendation, level II evidence).
• Routine testing to identify the cause of CAP in outpatients is optional (moderate recommendation, level III evidence).

TABLE
Clinical indications for more extensive diagnostic testing

CLINICAL INDICATION	RECOMMENDED DIAGNOSTIC TESTS
	BLOOD CULTURE	SPUTUM CULTURE	LEGIONELLA URINARY ANTIGEN TEST	PNEUMOCOCCAL URINARY ANTIGEN TEST
ICU admission*
Failed outpatient therapy
Cavitary infiltrates†
Leukopenia
Active alcohol abuse
Chronic severe liver disease
Severe lung disease
Asplenia
Anatomic or functional
Recent travel‡
Within past 2 weeks
Positive Legionella urinary antigen test			N/A
Positive Pneumococcal urinary antigen test				N/A
Pleural effusion**
Additional tests:
* Endotracheal aspirate if intubated, possibly bronchoscopy or nonbronchoscopic bronchoalveolar lavage.
† Fungal and tuberculosis cultures.
‡ Region/type of travel related to Legionella, Coccidioides, Hantavirus, B pseudomallei, avian influenza, SARS.
** Thoracentesis and pleural fluid cultures.
Adapted from Mandell et al.1

Empiric antibiotics

The recommendations of IDSA/ATS are generally for a class of antibiotics rather than a specific drug, unless noted.

4. Assess DRSP risk factors

Growth of drug-resistant Streptococcus pneumoniae (DRSP) necessitated a more extensive list of risk factors for DRSP. Other recommendations did not change.

Outpatient treatment

• Adults who were previously healthy and who do not have risk factors for DRSP CAP should be treated with either a macrolide (azithromycin, clarithromycin, or erythromycin) (strong recommendation, level I evidence) or doxycycline (weak recommendation; level III evidence).
• In the presence of comorbidities that increase the risk for DRSP, these antibiotics are appropriate: a respiratory fluoroquinolone (moxifloxacin, gemifloxacin, or levofloxacin [750 mg/day dose]) (strong recommendation, level I evidence); or β-lactam plus a macrolide: (high-dose amoxicillin [eg, 1 g 3x daily] [strong recommendation, level I evidence] or amoxicillin-clavulanate [2 g twice daily] is preferred; but alternatives include ceftriaxone, cefpodoxime, and cefuroxime [500 mg twice daily]. Doxycycline [level II evidence] is an alternative to the macrolide.)

Inpatient non-ICU treatment

• β-lactam plus a macrolide (strong recommendation, level I evidence) (cefotaxime, ceftriaxone, and ampicillin; ertapenem for selected patients; doxycycline [level III evidence] is an alternative to the macrolide.).
• A respiratory fluoroquinolone (strong recommendation, level I evidence) is the treatment of choice for penicillin-allergic patients.

5. Assess MRSA risk factors

Although similar to the prior ATS guidelines, the new guidelines have added specific risk factors for community-acquired methicillin-resistant S aureus (MRSA). This change reflects the increasing prevalence of community-acquired MRSA as an etiology for CAP.

The new guidelines state that the overwhelming majority of CAP pathogens will be adequately treated with the recommended empiric regimens. Exceptions are infections due to community-acquired methicillin-resistant S aureus and Pseudomonas aeruginosa.

ICU treatment

• A β-lactam (cefotaxime, ceftriaxone, or ampicillin-sulbactam) plus either azithromycin (strong recommendation, level II evidence) or a respiratory fluoroquinolone (strong recommendation, level I evidence).
• For penicillin-allergic patients, a respiratory fluoroquinolone and aztreonam are recommended.
• For Pseudomonas infection (see FIGURE for risk factors), use an anti-pneumococcal, antipseudomonal β-lactam (piperacillin-tazobactam, cefepime, imipenem, or meropenem) plus either ciprofloxacin or levofloxacin (750-mg dose) or the above β-lactam plus an aminoglycoside and azithromycin or the above β-lactam plus aminoglycoside and a respiratory fluoroquinolone (moderate recommendation, level II evidence). Fifteen days of therapy may be more effective in Pseudomonas CAP based on nosocomial infection data.
• For community-acquired methicillin-resistant S aureus infection (see FIGURE for risk factors), add vancomycin or linezolid (moderate recommendation, level III evidence).

Diagnostic testing is of high yield for patients with severe CAP requiring ICU admission, allowing for early de-escalation of empirical treatment if results are negative.

FIGURE
Treatment of community-acquired pneumonia*

Pathogen-directed therapy

6. Identify the pathogen

New guidelines recommend that, once the pathogen is identified by reliable microbiological methods, therapy should be directed towards that specific pathogen to prevent increased resistance in the community (moderate recommendation, level III evidence).

Influenza. Treatment within 48 hours of symptom onset with oseltamivir or zanamivir is recommended for influenza A (strong recommendation, level I evidence).

While these antimicrobials should not be used in uncomplicated influenza with symptoms for >48 hours (level I evidence), they may be used in hospitalized patients or influenza pneumonia to reduce viral shedding (moderate recommendation, level III evidence).

Time to first dose

7. Start treatment in the ED

Rather than designating a time window for starting treatment, the IDSA/ATS committee recommended that patients receive the first antibiotic dose in the Emergency Department (moderate recommendation, level III evidence).

This newly added statement contrasts with some current quality measures that grade hospitals according to whether antibiotics are started within a specific time frame.

Duration of antibiotics

8. Base duration on specific criteria

Reflecting evidence that shorter courses appear to be as effective as longer courses, the newer guidelines recommend discontinuation when the patient meets specific clinical criteria. Before discontinuing antibiotics, all patients with CAP should:

• Be treated for at least 5 days (level I evidence),
• Be afebrile for 48 to 72 hours, and
• Have no more than 1 of these criteria for clinical instability (moderate recommendation, level II evidence): temperature ≥37.8°C; heart rate ≥100 beats/min; respiratory rate ≥24 breaths/min; systolic blood pressure ≤90 mm Hg; arterial oxygen saturation ≤90% or pO₂ ≤60 mm Hg on room air; inability to maintain oral intake; altered mental status.

Switching from IV to oral

The guidelines, similar to the prior guidelines, recommend switching to oral therapy for hemodynamically stable patients who are clinically improving, able to ingest medications, and have a normally functioning gastrointestinal tract (strong recommendation, level II evidence).

Hospital discharge

The guidelines recommend that patients be discharged as soon as they are clinically stable and have a safe environment for continued care. Patients receiving oral therapy do not require inpatient observation (moderate recommendation, level II evidence). This is unchanged from prior recommendations.

The 2007 guidelines from the Infectious Diseases Society of America (IDSA)/American Thoracic Society (ATS)¹ are a blend of level-of-evidence strength and consensus opinion—a unified, evidence-based document. these new recommendations address prior discrepancies between the 2 specialties. We developed a CAP treatment algorithm based on the new advisory. (The following text includes levels of evidence.)

Site-of-care decisions

1. Let severity score be your guide

Based on evidence that physicians often hospitalize patients for CAP who could be managed as outpatients, the new guidelines recommend that we use an illness severity score (strong recommendation, level I evidence).

Previous guidelines advised only that we consider using a severity score.

• Use the validated Pneumonia Severity Index or the easier-to-use CURB-65. Patients with a CURB-65 score of 2 or more generally require hospitalization (moderate recommendation, level III evidence).
• Ability to reliably and safely take medications at home must also be taken into account (strong recommendation, level II evidence.)

2. Admit to ICU promptly if needed

The criteria for admission to the ICU is similar to the previous ATS guidelines, but the list of minor criteria is more extensive. This change reflects evidence demonstrating worse outcomes in patients whose transfer to the ICU was delayed. This new criteria has not been validated.

• Patients requiring vasopressors for blood pressure support or with hypoxemic respiratory failure should be admitted to the ICU—these are major criteria (strong recommendation, level II evidence).
• Patients with 3 or more minor criteria should also be directly admitted to the ICU (moderate recommendation, level II evidence).

3. Identify who needs more tests

In the wake of controversy about diagnostic testing recommendations, the new guidelines attempt to better identify patients who would benefit from further testing (TABLE).

• 12 indications. Prior ATS guidelines lacked specifics on required additional testing, but the new guidelines give 12 clinical indications for more extensive evaluation, and identify which tests are recommended for each indication (strong recommendation, level II evidence).
• Routine testing to identify the cause of CAP in outpatients is optional (moderate recommendation, level III evidence).

TABLE
Clinical indications for more extensive diagnostic testing

Empiric antibiotics

The recommendations of IDSA/ATS are generally for a class of antibiotics rather than a specific drug, unless noted.

4. Assess DRSP risk factors

Growth of drug-resistant Streptococcus pneumoniae (DRSP) necessitated a more extensive list of risk factors for DRSP. Other recommendations did not change.

Outpatient treatment

• Adults who were previously healthy and who do not have risk factors for DRSP CAP should be treated with either a macrolide (azithromycin, clarithromycin, or erythromycin) (strong recommendation, level I evidence) or doxycycline (weak recommendation; level III evidence).
• In the presence of comorbidities that increase the risk for DRSP, these antibiotics are appropriate: a respiratory fluoroquinolone (moxifloxacin, gemifloxacin, or levofloxacin [750 mg/day dose]) (strong recommendation, level I evidence); or β-lactam plus a macrolide: (high-dose amoxicillin [eg, 1 g 3x daily] [strong recommendation, level I evidence] or amoxicillin-clavulanate [2 g twice daily] is preferred; but alternatives include ceftriaxone, cefpodoxime, and cefuroxime [500 mg twice daily]. Doxycycline [level II evidence] is an alternative to the macrolide.)

Inpatient non-ICU treatment

• β-lactam plus a macrolide (strong recommendation, level I evidence) (cefotaxime, ceftriaxone, and ampicillin; ertapenem for selected patients; doxycycline [level III evidence] is an alternative to the macrolide.).
• A respiratory fluoroquinolone (strong recommendation, level I evidence) is the treatment of choice for penicillin-allergic patients.

5. Assess MRSA risk factors

Although similar to the prior ATS guidelines, the new guidelines have added specific risk factors for community-acquired methicillin-resistant S aureus (MRSA). This change reflects the increasing prevalence of community-acquired MRSA as an etiology for CAP.

The new guidelines state that the overwhelming majority of CAP pathogens will be adequately treated with the recommended empiric regimens. Exceptions are infections due to community-acquired methicillin-resistant S aureus and Pseudomonas aeruginosa.

ICU treatment

• A β-lactam (cefotaxime, ceftriaxone, or ampicillin-sulbactam) plus either azithromycin (strong recommendation, level II evidence) or a respiratory fluoroquinolone (strong recommendation, level I evidence).
• For penicillin-allergic patients, a respiratory fluoroquinolone and aztreonam are recommended.
• For Pseudomonas infection (see FIGURE for risk factors), use an anti-pneumococcal, antipseudomonal β-lactam (piperacillin-tazobactam, cefepime, imipenem, or meropenem) plus either ciprofloxacin or levofloxacin (750-mg dose) or the above β-lactam plus an aminoglycoside and azithromycin or the above β-lactam plus aminoglycoside and a respiratory fluoroquinolone (moderate recommendation, level II evidence). Fifteen days of therapy may be more effective in Pseudomonas CAP based on nosocomial infection data.
• For community-acquired methicillin-resistant S aureus infection (see FIGURE for risk factors), add vancomycin or linezolid (moderate recommendation, level III evidence).

Diagnostic testing is of high yield for patients with severe CAP requiring ICU admission, allowing for early de-escalation of empirical treatment if results are negative.

FIGURE
Treatment of community-acquired pneumonia*

Pathogen-directed therapy

6. Identify the pathogen

New guidelines recommend that, once the pathogen is identified by reliable microbiological methods, therapy should be directed towards that specific pathogen to prevent increased resistance in the community (moderate recommendation, level III evidence).

Influenza. Treatment within 48 hours of symptom onset with oseltamivir or zanamivir is recommended for influenza A (strong recommendation, level I evidence).

While these antimicrobials should not be used in uncomplicated influenza with symptoms for >48 hours (level I evidence), they may be used in hospitalized patients or influenza pneumonia to reduce viral shedding (moderate recommendation, level III evidence).

Time to first dose

7. Start treatment in the ED

Rather than designating a time window for starting treatment, the IDSA/ATS committee recommended that patients receive the first antibiotic dose in the Emergency Department (moderate recommendation, level III evidence).

This newly added statement contrasts with some current quality measures that grade hospitals according to whether antibiotics are started within a specific time frame.

Duration of antibiotics

8. Base duration on specific criteria

Reflecting evidence that shorter courses appear to be as effective as longer courses, the newer guidelines recommend discontinuation when the patient meets specific clinical criteria. Before discontinuing antibiotics, all patients with CAP should:

• Be treated for at least 5 days (level I evidence),
• Be afebrile for 48 to 72 hours, and
• Have no more than 1 of these criteria for clinical instability (moderate recommendation, level II evidence): temperature ≥37.8°C; heart rate ≥100 beats/min; respiratory rate ≥24 breaths/min; systolic blood pressure ≤90 mm Hg; arterial oxygen saturation ≤90% or pO₂ ≤60 mm Hg on room air; inability to maintain oral intake; altered mental status.

Switching from IV to oral

The guidelines, similar to the prior guidelines, recommend switching to oral therapy for hemodynamically stable patients who are clinically improving, able to ingest medications, and have a normally functioning gastrointestinal tract (strong recommendation, level II evidence).

Hospital discharge

The guidelines recommend that patients be discharged as soon as they are clinically stable and have a safe environment for continued care. Patients receiving oral therapy do not require inpatient observation (moderate recommendation, level II evidence). This is unchanged from prior recommendations.

The 2007 guidelines from the Infectious Diseases Society of America (IDSA)/American Thoracic Society (ATS)¹ are a blend of level-of-evidence strength and consensus opinion—a unified, evidence-based document. these new recommendations address prior discrepancies between the 2 specialties. We developed a CAP treatment algorithm based on the new advisory. (The following text includes levels of evidence.)

Site-of-care decisions

1. Let severity score be your guide

Based on evidence that physicians often hospitalize patients for CAP who could be managed as outpatients, the new guidelines recommend that we use an illness severity score (strong recommendation, level I evidence).

Previous guidelines advised only that we consider using a severity score.

• Use the validated Pneumonia Severity Index or the easier-to-use CURB-65. Patients with a CURB-65 score of 2 or more generally require hospitalization (moderate recommendation, level III evidence).
• Ability to reliably and safely take medications at home must also be taken into account (strong recommendation, level II evidence.)

2. Admit to ICU promptly if needed

The criteria for admission to the ICU is similar to the previous ATS guidelines, but the list of minor criteria is more extensive. This change reflects evidence demonstrating worse outcomes in patients whose transfer to the ICU was delayed. This new criteria has not been validated.

• Patients requiring vasopressors for blood pressure support or with hypoxemic respiratory failure should be admitted to the ICU—these are major criteria (strong recommendation, level II evidence).
• Patients with 3 or more minor criteria should also be directly admitted to the ICU (moderate recommendation, level II evidence).

3. Identify who needs more tests

In the wake of controversy about diagnostic testing recommendations, the new guidelines attempt to better identify patients who would benefit from further testing (TABLE).

• 12 indications. Prior ATS guidelines lacked specifics on required additional testing, but the new guidelines give 12 clinical indications for more extensive evaluation, and identify which tests are recommended for each indication (strong recommendation, level II evidence).
• Routine testing to identify the cause of CAP in outpatients is optional (moderate recommendation, level III evidence).

TABLE
Clinical indications for more extensive diagnostic testing

Empiric antibiotics

The recommendations of IDSA/ATS are generally for a class of antibiotics rather than a specific drug, unless noted.

4. Assess DRSP risk factors

Growth of drug-resistant Streptococcus pneumoniae (DRSP) necessitated a more extensive list of risk factors for DRSP. Other recommendations did not change.

Outpatient treatment

• Adults who were previously healthy and who do not have risk factors for DRSP CAP should be treated with either a macrolide (azithromycin, clarithromycin, or erythromycin) (strong recommendation, level I evidence) or doxycycline (weak recommendation; level III evidence).
• In the presence of comorbidities that increase the risk for DRSP, these antibiotics are appropriate: a respiratory fluoroquinolone (moxifloxacin, gemifloxacin, or levofloxacin [750 mg/day dose]) (strong recommendation, level I evidence); or β-lactam plus a macrolide: (high-dose amoxicillin [eg, 1 g 3x daily] [strong recommendation, level I evidence] or amoxicillin-clavulanate [2 g twice daily] is preferred; but alternatives include ceftriaxone, cefpodoxime, and cefuroxime [500 mg twice daily]. Doxycycline [level II evidence] is an alternative to the macrolide.)

Inpatient non-ICU treatment

• β-lactam plus a macrolide (strong recommendation, level I evidence) (cefotaxime, ceftriaxone, and ampicillin; ertapenem for selected patients; doxycycline [level III evidence] is an alternative to the macrolide.).
• A respiratory fluoroquinolone (strong recommendation, level I evidence) is the treatment of choice for penicillin-allergic patients.

5. Assess MRSA risk factors

Although similar to the prior ATS guidelines, the new guidelines have added specific risk factors for community-acquired methicillin-resistant S aureus (MRSA). This change reflects the increasing prevalence of community-acquired MRSA as an etiology for CAP.

The new guidelines state that the overwhelming majority of CAP pathogens will be adequately treated with the recommended empiric regimens. Exceptions are infections due to community-acquired methicillin-resistant S aureus and Pseudomonas aeruginosa.

ICU treatment

• A β-lactam (cefotaxime, ceftriaxone, or ampicillin-sulbactam) plus either azithromycin (strong recommendation, level II evidence) or a respiratory fluoroquinolone (strong recommendation, level I evidence).
• For penicillin-allergic patients, a respiratory fluoroquinolone and aztreonam are recommended.
• For Pseudomonas infection (see FIGURE for risk factors), use an anti-pneumococcal, antipseudomonal β-lactam (piperacillin-tazobactam, cefepime, imipenem, or meropenem) plus either ciprofloxacin or levofloxacin (750-mg dose) or the above β-lactam plus an aminoglycoside and azithromycin or the above β-lactam plus aminoglycoside and a respiratory fluoroquinolone (moderate recommendation, level II evidence). Fifteen days of therapy may be more effective in Pseudomonas CAP based on nosocomial infection data.
• For community-acquired methicillin-resistant S aureus infection (see FIGURE for risk factors), add vancomycin or linezolid (moderate recommendation, level III evidence).

Diagnostic testing is of high yield for patients with severe CAP requiring ICU admission, allowing for early de-escalation of empirical treatment if results are negative.

FIGURE
Treatment of community-acquired pneumonia*

Pathogen-directed therapy

6. Identify the pathogen

New guidelines recommend that, once the pathogen is identified by reliable microbiological methods, therapy should be directed towards that specific pathogen to prevent increased resistance in the community (moderate recommendation, level III evidence).

Influenza. Treatment within 48 hours of symptom onset with oseltamivir or zanamivir is recommended for influenza A (strong recommendation, level I evidence).

While these antimicrobials should not be used in uncomplicated influenza with symptoms for >48 hours (level I evidence), they may be used in hospitalized patients or influenza pneumonia to reduce viral shedding (moderate recommendation, level III evidence).

Time to first dose

7. Start treatment in the ED

Rather than designating a time window for starting treatment, the IDSA/ATS committee recommended that patients receive the first antibiotic dose in the Emergency Department (moderate recommendation, level III evidence).

This newly added statement contrasts with some current quality measures that grade hospitals according to whether antibiotics are started within a specific time frame.

Duration of antibiotics

8. Base duration on specific criteria

Reflecting evidence that shorter courses appear to be as effective as longer courses, the newer guidelines recommend discontinuation when the patient meets specific clinical criteria. Before discontinuing antibiotics, all patients with CAP should:

• Be treated for at least 5 days (level I evidence),
• Be afebrile for 48 to 72 hours, and
• Have no more than 1 of these criteria for clinical instability (moderate recommendation, level II evidence): temperature ≥37.8°C; heart rate ≥100 beats/min; respiratory rate ≥24 breaths/min; systolic blood pressure ≤90 mm Hg; arterial oxygen saturation ≤90% or pO₂ ≤60 mm Hg on room air; inability to maintain oral intake; altered mental status.

Switching from IV to oral

The guidelines, similar to the prior guidelines, recommend switching to oral therapy for hemodynamically stable patients who are clinically improving, able to ingest medications, and have a normally functioning gastrointestinal tract (strong recommendation, level II evidence).

Hospital discharge

The guidelines recommend that patients be discharged as soon as they are clinically stable and have a safe environment for continued care. Patients receiving oral therapy do not require inpatient observation (moderate recommendation, level II evidence). This is unchanged from prior recommendations.