Home > Health & Research > Health Education Campaigns & Programs > Cochrane Neonatal Review > Erythromycin for the prevention of chronic lung disease in intubated preterm infants at risk for, or colonized or infected with Ureaplasma urealyticum

Erythromycin for the prevention of chronic lung disease in intubated preterm infants at risk for, or colonized or infected with Ureaplasma urealyticum

Skip sharing on social media links
Share this:

Authors

Mabanta CG, Pryhuber GS, Weinberg GA, Phelps DL

Background - Methods - Results - Characteristics of Included Studies - References - Data Tables & Graphs


Dates

Date edited: 25/08/2003
Date of last substantive update: 09/06/2003
Date of last minor update: 02/02/2003
Date next stage expected / /
Protocol first published: Issue 3, 2002
Review first published: Issue 4, 2003

Contact reviewer

Dr Carmelita G Mabanta

Fellow
Pediatrics-Neonatology
Strong Children's Research Center, University of Rochester
601 Elmwood Avenue Box 651
Rochester
New York USA
14642
Telephone 1: 585 275 5884
Facsimile: 585 461 3614

E-mail: carmelita_mabanta@urmc.rochester.edu

Secondary contact person's name: Dale L. Phelps, M.D.

Contribution of reviewers

Mabanta CG:
Contact person
Submission of title to Neonatal Review Group
Development of protocol
Literature search
Screening of possibly relevant trials
Evaluation of papers and extraction of data
Entering data in RevMan
Communicating with authors for clarification, additional data, ongoing studies
Writing of protocol and full review

Pryhuber GS:
Help in development of the protocol
Evaluation of papers and extraction of data
Providing recommendations for the review; editing final review

Weinberg GA:
Help in development of the protocol
Evaluation of papers and extraction of data
Providing recommendations for the review; editing final review

Phelps DL:
Introduction and guidelines in writing a Cochrane Review
Help in the development of the protocol
Evaluation of papers and extraction of data
Help in entering data in RevMan
Help in writing of the protocol and the review
Providing recommendations for the review; editing final review

Sources of Support

Internal sources of support

Division of Neonatology, Strong Children's Research Center, University of Rochester, Rochester, New York, USA

External sources of support

  • None noted.

What's new

History

Date review re-formatted: / /
Date new studies sought but none found: 09/06/2003
Date new studies found but not yet included/excluded: 15/02/2003
Date new studies found and included/excluded: / /
Date reviewers' conclusions section amended: / /
Date comment/criticism added: / /
Date response to comment/criticisms added: / /

Synopsis

More evidence needed on using antibiotics to try and prevent chronic lung disease in preterm babies testing positive for U urealyticum.

Chronic lung disease (CLD) is a major health risk often faced by preterm babies on ventilation (machine-assisted breathing). Ureaplasma urealyticum is a common type of infection that may be one of the causes of CLD, although this has not been proven. Erythromycin is an antibiotic used for other infections caused by U urealyticum. When a baby tests positive for U urealyticum, erythromycin has been tried as a preventive measure. The review found that there is not enough evidence from trials to show whether or not testing preterm babies for U urealyticum and using erythromycin can prevent CLD or mortality.

Abstract

Background

Controversy exists over whether or not Ureaplasma urealyticum colonization or infection of the respiratory tract contributes to the severity of chronic lung disease (CLD), a major cause of morbidity and mortality in preterm infants.

Objectives

To evaluate the efficacy and safety of prophylactic or therapeutic erythromycin in preventing chronic lung disease in intubated preterm infants with unknown U. urealyticum status or proven positivity.

Search strategy

Searches were done of MEDLINE (1966-June 9, 2003), EMBASE (1980-May 5, 2003), The Cochrane Central Register of Controlled Trials (CENTRAL, The Cochrane Library, Issue 4, 2002), previous reviews including cross-references, and abstracts of conference proceedings (Pediatric Academic Societies 2000-2003, American Thoracic Society 2001-2003). There were no language restrictions. Expert informants were contacted.

Selection criteria

Randomized or quasi-randomized studies comparing either prophylactic or therapeutic administration of oral or intravenous erythromycin (regardless of dose and duration) versus no treatment or placebo among intubated preterm infants < 37 weeks and < 2500 grams with either unknown U. urealyticum status or proven positivity by culture or polymerase chain reaction.

Data collection & analysis

Data were extracted by all of the authors independently and differences were resolved by consensus. Treatment effects for categorical outcomes were expressed as relative risk, with 95% confidence intervals.

Main results

Two small controlled studies, both involving intubated babies < 30 weeks gestation, were eligible for inclusion. Lyon 1998 tested prophylactic erythromycin in babies whose U. urealyticum status was unknown at the time of initiation of treatment. Jonsson 1998 tested erythromycin in babies known to be culture positive for U. urealyticum. Neither trial showed a statistically significant effect of erythromycin on CLD, death or the combined outcome CLDor death. Because the two studies differed importantly in their design, theresults were not combined in meta-analyses. No adverse effects of a 7-10day course of erythromycin were reported in either study.

Reviewers' conclusions

Current evidence does not demonstrate a reduction in CLD/death when intubated preterm infants are treated with erythromycin prophylactically before U. urealyticumculture/PCR results are known or when Ureaplasma colonized, intubated preterminfants are treated with erythromycin. However, a true benefit could easilyhave been missed with the small sample sizes in the two eligible studies. The studies were greatly underpowered to detect uncommon adverse effects such as pyloric stenosis. Additional controlled trials are required to determine whether antibiotic therapy of Ureaplasma reduces CLD and/or death in intubated preterm infants.

[top]

Background

Controversy exists over whether or not Ureaplasma urealyticum colonization or infection of the respiratory tract contributes to the severity of chronic lung disease (CLD), a major cause of morbidity and mortality in preterm infants.

U. urealyticum is an atypical bacterial organism that lacks a rigid cell wall, therefore making it resistant to beta-lactam antibiotics. Mucosal surfaces, particularly the respiratory and genital tracts, are commonly colonized (Tully 1993; Cassell 2001), but symptomatic infection is relatively uncommon.

In adults, U. urealyticum has been associated with urethritis in males, urinary calculi (Grenabo 1988; Cassell 2001) and, in immunocompromised patients, arthritis (Forgacs 1993) and osteomyelitis (Mohiuddin 1991). The organism can be found in the vagina or cervix of 40-80% of sexually mature, asymptomatic women (McCormack 1986a; Cassell 2001). Colonization is linked to younger age, lower socioeconomic status, sexual activity with multiple partners, oralcontraceptive use and black ethnicity (McCormack 1986a; McCormack 1986b). The presence of U. urealyticum in the amniotic fluid and placenta has been associated withadverse pregnancy outcomes such as chorioamnionitis, preterm delivery, spontaneousabortion and stillbirths (Cassell 1991). Transmissionof U. urealyticum to the fetus may occur in utero via contamination of theamniotic fluid with organisms ascending from the vagina or during deliveryby passage through a colonized birth canal (Dinsmoor1989). Vertical transmission results in colonization rates from 46-89%in preterm infants and is highest in infants with birthweights less than1000 grams (Sanchez 1990). Rates of transmissionare somewhat lower (44-55%) in full term infants (Sanchez 1987; Syrogiannopoulos '90). Neonatal morbidity possibly associated with U. urealyticum includes neonatal pneumonia (Waites 1989) and meningitis (Garland 1987). U. urealyticum has also been implicatedas contributing to the incidence and severity of chronic lung disease ininfants born prematurely. However, whether U. urealyticum infection is causativeof these conditions is unclear.

The lungs of preterm infants are prone to injury leading to CLD. Lung injury may be due to several factors such as use of ventilators, oxygen supplementation and infection. Neonatal CLD was first described as a spectrum of clinical, radiological and pathologic findings in infants requiring mechanical ventilation for respiratory distress syndrome (Northway 1967). The term bronchopulmonary dysplasia (BPD) was used to describe the abnormal pulmonary growth, development and fibrosis evident histologically in infants dying of CLD. Later, the clinical criteria for the diagnosis of BPD in survivorswere defined as a requirement for oxygen beyond one month of age plus characteristicx-ray findings. Since the initial description of BPD by Northway, however, the clinical picture has changed. Very premature infants survive more oftenand the histopathology observed demonstrates fewer airway abnormalities andmore alveolar fibrosis and failure of alveolarization. The need for supplementaloxygen at 28 days of life was a good predictor of abnormal pulmonary findingsin infants born equal to or greater than 30 weeks of gestation but becameless useful as a diagnostic criterion as the gestational age of survivorsdecreased. For the younger infants the requirement for additional oxygenat 36 weeks postmenstrual age (PMA) is a better predictor of abnormal pulmonaryoutcomes in the first two years of life (Shennan1988). It has been proposed that U. urealyticum infection or colonizationat birth results in an increased risk of CLD.

There are several mechanisms proposed for how U. urealyticum may contribute to the severity of CLD in preterm infants. Neonatal fibroblasts release significantlevels of IL-6 and IL-8 in direct response to active infection with U. urealyticumeven without the added insult of hyperoxia (Stancombe1993). Thus, U. urealyticum could initiate or escalate an inflammatoryresponse in the lungs by inducing these cytokines. Increased numbers ofpolymorphonuclear cells, consistent with reactive inflammation, were detectedin endotracheal aspirates from preterm infants colonized with U. urealyticumon the second day of life (Payne 1993). Also, clinical or subclinical Ureaplasma pneumonia may increase oxygen requirementsand increase the severity of CLD resulting from oxygen toxicity (Cassell 1988a). Despite the potential for U.urealyticum to cause lung inflammation and injury, published studies do notagree on the role of the organism in the development of CLD. Differencesin the patient populations, culture methods and the definition of CLD utilizedin the studies contribute to the disparities in reported results.

Several studies have implicated colonization with U. urealyticum in the development of CLD in preterm infants (Cassell 1988b; Sanchez 1988; Wang 1988; Horowitz 1992; Crouse 1993; Payne 1993; Iles 1996; Perzigian 1998; Hannaford 1999). The majority ofthe studies used a definition of CLD as "the need for supplemental oxygen at 28 days of life." There were differences in defining colonization because of various culture techniques. All of the studies cultured nasopharyngeal aspirates and/or endotracheal aspirates for U. urealyticum. In addition, infant specimens cultured for U. urealyticum included blood (Cassell 1988b) and swabs from the throat, eye, vagina and rectum (Sanchez 1988). Endotracheal aspirates were found to be the most significant marker of colonization in the respiratory tract (Cassell 1988b). U. urealyticum was isolated from endotracheal aspirates in pure culture from 85% of preterm infants and recovered at the same time from blood culture in26% of these infants (Cassell 1988b). Ina recent meta-analysis (Wang 1995), the relative risk (RR) of CLD in colonized compared to non-colonized infants was 1.72 (95% confidence interval, 1.50 - 1.96).

Other studies have suggested that U. urealyticum colonization has no association with the development of CLD (Heggie 2001; Ollikainen 2001; van Waarde 1997; Saxen 1993; Smyth 1993). Ollikainen found no association between U. urealyticum colonization and CLDboth at 28 days after birth and 36 weeks PMA. Heggie found no association between colonization with two distinct species of U. urealyticum (U. urealyticum and U. parvum), which have possible differences in pathogenicity in preterm infants, and CLD at 36 weeks PMA. Logistic regression analysis with CLD asthe dependent variable, and gestational age, presence of U. urealyticum (van Waarde 1997; Smyth 1993) and birth weight (Smyth 1993) as the independent variables, showed that CLD was significantly related to decreasing gestational age but not tothe presence of a positive U. urealyticum culture (van Waarde 1997; Smyth 1993 ) or birth weight (Smyth 1993).

One difference that may contribute to the disagreement between studies concerning the association of U. urealyticum with the development of CLD isthe frequency of maternal colonization in the populations studied. The studiesimplicating U. urealyticum in the development of CLD (Sanchez 1988; Cassell 1988b; Crouse 1993 ) were conducted with inner city populations with high rates of colonization. In contrast, the studies that did not find a relation to CLD (Saxen 1993; Ollikainen 2001; Smyth 1993 ) were conducted in Finnish and British populations utilizing populations with low maternal genital colonization rates. This suggests that U. urealyticum, if not causal of CLD itself, may be a surrogate marker of risk factors affecting lower socio-economic classes, secondarily affecting CLD.

Differing study results concerning the association between U. urealyticum and CLD have led to uncertainty about the possible efficacy of treating infections/colonizations of the respiratory tract in the neonate. Although use of erythromycin to treat U. urealyticum positive cultures in preterm babies is common, no studies have demonstrated the efficacy of such treatment. A causative relationship of U. urealyticum to CLD in preterm infants would be supported if treatment of the organism reduced or prevented the lung disease. Erythromycin has beenthe drug of choice for U. urealyticum infections in the respiratory tract.A recent review (Buhrer 2001), which was nota meta analysis, examined one non-randomized and two randomized controlled trials using erythromycin for U. urealyticum in preterm infants. The authors concluded that antibiotic treatment failed to reduce chronic lung disease. This Cochrane Review will evaluate studies testing erythromycin therapy fordecreasing the incidence and severity of chronic lung disease in preterm infantsintubated after birth. Side effects of erythromycin will also be evaluated.

Objectives

To evaluate the efficacy and safety of erythromycin in preventing chronic lung disease in intubated preterm infants, when used:

  1. prophylactically in infants with unknown U. urealyticum status
  2. therapeutically in infants known to be U. urealyticum positive

This objective has been modified from the previously published protocolto separate prophylactic treatment from treatment of known positive infants.

Criteria for considering studies for this review

Types of studies

Randomized or quasi-randomized studies.

Types of participants

Intubated preterm infants < 37 weeks and < 2500 grams with unknown U. urealyticum status or with proven U. urealyticum positivity [by culture or polymerase chain reaction (PCR)]. There is no restriction with regards to the site of origin of the samples used for culture or polymerase chain reaction (PCR) tests.

Types of interventions

Erythromycin administered orally or intravenously (regardless of dose and duration) for prophylactic or therapeutic treatment of U. urealyticum, vs. no treatment or placebo.

Types of outcome measures

The primary outcome measure is the occurrence of death and/or CLD, defined as a requirement for supplemental oxygen at 36 weeks PMA.

Secondary outcome measures are the days on mechanical ventilation, days on supplemental oxygen, retinopathy of prematurity, necrotizing enterocolitis and the possible side effects of erythromycin treatment including: gastrointestinal disturbances (e.g. diarrhea, vomiting), cardiac tachyarrhythmias, ability of erythromycin to decrease the metabolism of theophylline causing toxicity (e.g. hyperglycemia, hyperreflexia, jitteriness, seizures), or pyloric stenosis.

Search strategy for identification of studies

Search of literature in:

MEDLINE from 1966 - June 9, 2003
EMBASE from 1980 - May 5, 2003
Cochrane Central Register of Controlled Trials (CENTRAL, The Cochrane Library, Issue 4, 2002)
Previous reviews including cross-references, abstracts (Pediatric Academic Societies 2000-2003 and American Thoracic Society meetings 2001-2003), expert informants
Subjects were limited to the human population. There were no language restrictions.

Search terms included these key words and/or MeSH headings: [preterm infant and limit to human] and [bronchopulmonary dysplasia or chronic lung disease] and [Ureaplasma or erythromycin]

[top]

Methods of the review

The standard methods of the Cochrane Neonatal Review Group were used. Titles and abstracts of reports obtained from the databases were initially assessed by one reviewer (CM). The entire article was obtained for studies that were of possible interest and these were assessed independently by the four reviewers for relevance. When needed, authors of abstracts or full reportswere contacted for clarification.

The validity of the eligible studies was assessed independently by the four reviewers. Differences were resolved by consensus. Each eligible study was evaluated for:

  1. Selection bias (blinding of randomization)
  2. Performance bias (blinding of intervention)
  3. Attrition bias (complete follow-up)
  4. Detection bias (blinding of outcome measurement)

Outcome data for the included studies were abstracted independently by the four reviewers. Differences were resolved by consensus. Measures of treatment effect for categorical outcomes were relative risk (RR), risk difference (RD), and number needed to treat (NNT). 95% confidence intervals were used. A fixed effect model was planned for meta-analysis. Tests of heterogeneity were not done as the identified studies could not be combined. Sensitivity analyses were not done.

Description of studies

One hundred forty titles (full reports or abstracts) were screened and, of these, seventeen possibly relevant studies were selected for review by all authors. Two controlled trials (Lyon 1998; Jonsson 1998) were judged to be eligible for inclusion and are described below. Additional data on these two studies wereobtained by personal communication with these authors. Fifteen studies wereexcluded (see the details in the Table of Excluded Studies). One study (Okura 1998) is awaiting assessment.

Included studies (see table, Characteristics of Included Studies, for further details).

Lyon 1998 - The objective of the study was todetermine if erythromycin given prophylactically from birth in preterm infantsreduced the inflammatory response and the incidence and severity of CLD.Intubated infants equal to or less than 30 weeks gestation (n=75) were randomized(used sealed envelopes containing random number table allocation) at birthto receive erythromycin (15/mg/kg/dose) three times a day for 7 days or not.The primary outcome was to compare the levels of interleukin-1 beta (IL-1beta), interleukin-8 (IL-8) and tumor necrosis factor alpha (TNF-alpha) betweenthe treatment and control groups. Culture and PCR for U. urealyticum wereobtained from tracheobronchial secretions on admission. Nine infants (13%)proved to be positive for U. urealyticum of which three were in the treatmentgroup and six were in the control group. The clinical staff in the NICU didnot know the culture or PCR results until after the study was completed.CLD was defined as the need for supplemental oxygen at 36 weeks PMA. Alveolararterial oxygen difference (AaDO2) at 36 weeks PMA, which correlates withthe degree of pulmonary dysfunction at 1 year of age in babies with CLD, was measured to assess the severity of lung disease (Iles 1997).

Jonsson 1998 - The aim of the study was to investigate U. urealyticum colonization and the effect of erythromycin treatment in culture positive infants less than 30 weeks gestation. Other risk factors influencing respiratory morbidity and the need for supplemental oxygen were also evaluated. One hundred fifty-five intubated infants < 30 weeks gestation were screened in the study. Cultures for U. urealyticum from endotracheal aspirates and nasopharyngeal swabs were taken as soon aspossible after birth. 29/155 (19%) of infants were culture positive. Onecolonized infant died due to severe IVH at day 3, was not allocated to treatmentor control and was, therefore, not included in the trial (personal communication).As soon as culture results were known (an average of 7 days after birth), every other colonized infant was treated with erythromycin (n=14). The 14treated babies received a 10-day course of erythromycin 40 mg/kg/day orally(PO) n=10, or, if they did not tolerate oral feeds, intravenously (IV) n=4.The remaining 14 infants did not receive erythromycin or placebo.

Methodological quality of included studies

Lyon 1998:

Blinding of randomization - Yes (used sealed envelopes)
Blinding of intervention - No (placebo not used)
Complete follow-up - Yes
Blinding of outcome measurement - Yes

Jonsson 1998:

Blinding of randomization - No (quasi-randomization by alternate assignment)
Blinding of intervention - No (placebo not used)
Complete follow-up - Yes
Blinding of outcome measurement - No

[top]

Results

PROPHYLACTIC ERYTHROMYCIN VS NO TREATMENT OR PLACEBO IN BABIES AT RISK FOR U. UREALYTICUM COLONIZATION OR INFECTION (Lyon 1998)

  1. CLD
    CLD occurred in 13/34 treated babies and 11/41 controls. There was no significant effect of erythromycin prophylaxis on CLD [RR=1.40, 95% CI: 0.72, 2.70]. Since the effect may have differed among those who subsequently turned out to be culture/PCR positive, we also looked at the subgroups of positive and negative infants. There was no significant effect in either subgroup:
    1. Culture/PCR Positive - 1/3 infants had CLD in the treatment group and 1/6 infants in the control group [RR=2.00, 95% CI: 0.18, 22.06].
    2. Culture/PCR Negative - 12/31 infants had CLD in the treatment group and 10/35 in the control group [RR=1.35, 95% CI: 0.68, 2.69].
  2. Death
    Deaths occurred in 4/34 treated infants, and 8/41 controls. There was no significant effect of erythromycin either in all infants enrolled [RR=0.64, 95% CI: 0.22, 1.83] or in either subgroup:
    1. Culture/PCR Positive - 0/3 infants died in the treatment group and 1/6 died in the control group [RR=0.58, 95% CI: 0.03, 11.21].
    2. Culture/PCR Negative - 4/31 infants died in the treatment group and 7/35 infants died in the control group [RR=0.65, 95% CI: 0.21, 2.00].
  3. Death or CLD
    Combining Death or CLD is necessary because these are competing outcomes. Among the treated infants, 17/34 vs 19/41 control infants experienced death or CLD. There was no significant effect of erythromycin either in all infants enrolled [RR=1.06. 95% CI: 0.66, 1.69] or in either subgroup:
    1. Culture/PCR Positive - 1/3 infants died or had CLD in the treatment group and 2/6 infants died or had CLD in the control group [RR=1.00, 95% CI: 0.14, 7.10].
    2. Culture/PCR Negative - 16/31 infants died or had CLD in the treatment group and 17/35 infants died or had CLD in the control group [RR=1.06, 95% CI: 0.66, 1.72].
  4. Side effects of erythromycin
    There were no recognized side effects of erythromycin observed in the treated infants (personal communication).
  5. Number of days on mechanical ventilation
    In Lyon's study, these parameters were not reported by treatment group. However, among all culture positive infants, the median number of days on mechanical ventilation was 14 days (range 2-80 days) and in the culture negative group, the infants were ventilated for a median of 6 days (range 1-63 days).
  6. Number of days on supplemental oxygen
    Not reported
  7. Retinopathy of prematurity
    Not reported
  8. Necrotizing enterocolitis
    Not reported
  9. Effects on inflammatory responses were reported in this study, but are not included in this review

THERAPEUTIC ERYTHROMYCIN VS NO TREATMENT OR PLACEBO IN U. UREALYTICUM COLONIZEDOR INFECTED PRETERM INFANTS (Jonsson 1998)

  1. CLD
    9/14 infants had CLD in the treatment group versus 10/14 infants in the control group. Erythromycin had no significant effect [RR=0.90, 95% CI: 0.54, 1.50].
  2. Death
    There were no deaths in either group.
  3. Death or CLD
    9/14 infants had CLD or died in the treatment group versus 10/14 infants in the control group. Erythromycin had no significant effect [RR=0.90, 95% CI: 0.54, 1.50].
  4. Side effects of erythromycin
    There were no recognized side effects in the treated infants (personal communication).
  5. Number of days on mechanical ventilation
    The treated, culture-positive infants were ventilated for 9 days (median; range 1-47 days) and CPAP was used for a further 41 days (median; range 25-60 days). The infants who were not treated and culture-positive were ventilated for 6 days (median; range 1-29 days) and CPAP was used for a further 52 days (median; range 23-130 days).
    1. Among all of the culture-positive infants, ventilation was used for 6.5 days (median; range 1-47 days) and CPAP was used for a further 42 days (median; range 23-130 days). Among the culture-negative infants, ventilation was used for 7 days (median; range 1-420 days) and CPAP was used for a further 30 days (median; range 1-159 days).
  6. Number of days on supplemental oxygen
    The treated, culture-positive infants required supplemental oxygen for 82 days (median; range 37-180 days) and the infants who were not treated and culture-positive needed supplemental oxygen for 74 days (median; 48-180 days).
    1. Comparing the culture-positive infants with the culture-negative infants, culture-positive infants required supplemental oxygen for 80 days (median; range 37-180 days) while culture-negative infants required supplemental oxygen for 67 days (median; range 2-315 days).
  7. Retinopathy of prematurity
    Not reported
  8. Necrotizing enterocolitis
    Not reported

Discussion

The results of this review found no evidence that erythromycin decreased the incidence or severity of CLD in intubated preterm infants, whether used prophylactically or in infants known to be colonized or infected with U. urealyticum.Due to the small sample sizes, even a large benefit could have been missedby these studies. The means of detection of colonization (culture and/orPCR) and timing of treatment in both studies varied. Jonsson treated thecolonized infants after the culture results were known (mean=7 days) whileLyon treated infants at birth with erythromycin prior to the results of thecultures or PCR. We cannot combine the results of the two identified studiesbecause they are different in the populations and the timing of treatment. Awaiting the results of culture delayed treatment by about 7 days in one study, while the prophylactic study was able to treat from birth. Therefore, we did a separate analysis for infants who were given erythromycin prophylactically in contrast to therapeutically. The infants were divided into subgroups inLyon's study to see if there was a differential effect in those who were U.urealyticum positive

Although routine cultures for Ureaplasma are not done in pregnant women, it would be interesting to know the mothers' colonization status. Transmission of the organism may occur in utero or during delivery; therefore, early treatment of a culture positive mother might decrease the risk of colonization or infection in the fetus or infant. In a Cochrane Review, erythromycin use in women whohave preterm premature rupture of membranes showed a significant reduction in the number of babies born within 48 hours of membrane rupture [RR 0.84, 95% CI 0.76, 0.93], requiring oxygen therapy [RR 0.87, 95% CI 0.78, 0.98] and with a positive blood culture for any organism [RR 0.70 CI 0.52, 0.94] (Kenyon 2002).

Results of cultures for Ureaplasma can take a long time and the benefits of early treatment, if any, may be missed if treatment is delayed. PCR is equal to, if not more sensitive than culture for detection of Ureaplasma andgives more rapid results [ < 24 hours] (Blanchard1993). In a study where PCR was done in 100 endotracheal aspirates from54 intubated, preterm babies, PCR showed a sensitivity of 93% (Cunliffe 1996). Use of PCR could achieve promptdiagnosis and permit trials of early treatment in colonized mothers and/orinfants. However, cultures will still be important to determine drug sensitivities.

Side effects of erythromycin were not systematically reported in the included trials, or in other studies detected but excluded. Potential toxicity of the antibiotic is an important consideration especially if infants are to be treated prophylactically, prior to knowing the culture results. Since vertical transmission results in colonization rates among preterm infants of 46-89% (Sanchez 1990), a significant number of non-infected infants would be treated if prophylaxis were used, even if restricted to the infants of Ureaplasma positive mothers. Erythromycin use for pertussis in newborns has been associated with an increased risk of infantile hypertrophic pyloric stenosis, which raises concerns for the safety of this drug if it were to be used in a prophylactic manner. In a study of prophylactic erythromycin to prevent pertussis, the rate of pyloric stenosis was 32.3/1000 liveborn infants. This was about 7-fold higher than the rate of 4.7/1000 observed in infants born in the same hospital in the prior two years [RR=6.8; 95% CI 3.0-15.7] (Honein 1999).

Reviewers' conclusions

Implications for practice

Based on current evidence from controlled trials, we cannot recommend theroutine use of either prophylactic or therapeutic erythromycin for the purposeof reducing CLD secondary to U. urealyticum in intubated preterm infants.No evidence of benefit of treatment has been shown and the incidence ofside effects from the antibiotic is unknown. The decision to treat an individualpatient, however, is based on clinical status as well as culture results, and individual decisions may thus vary since the data lack the power to ruleout a true benefit. The fact that the rate of side effects of erythromycin in preterm infants is unknown cannot be overlooked.

Implications for research

Additional controlled clinical trials are required to determine whether antibiotic therapy of Ureaplasma reduces CLD/death in preterm infants.

Methods of detection of colonization and timing of treatment are important factors to consider in the design of trials to study efficacy of treatment of U. urealyticum in preventing CLD. Studies are needed to test 1) treatment of proven U. urealyticum colonization/infection, and 2) prophylaxis in high risk infants with unknown U. urealyticum status. Rapid PCR results from eitherthe mother or the infant could be used to identify groups of subjects whoare most likely to benefit from early treatment. Given the suggestions thatin utero fetal pulmonary inflammation may predispose to CLD, further considerationshould be given to randomized treatment of women who are carriers of U. urealyticum, or who are documented to have the organism in amniotic fluid, with follow-upnot only of effect on risk of preterm delivery but also for sequelae, includingCLD, in their newborns. It should also be considered that treatment of U.urealyticum infection/colonization of premature lungs may not affect themultifactorial disease that is CLD but may reduce the severity of the disease.Severity of CLD as well as its occurrence should be considered as studyoutcomes. All trials should also report other common neonatal morbiditiessuch as NEC and IVH, and should specifically look for and report potentialside effects of erythromycin.

Acknowledgements

Leandro Cordero, M.D., Baldvin Jonsson, M.D. and Andy Lyon, M.D. for providing unpublished data.

Gail Cassell, M.D., for providing references.

Luis Rodriguez, M.D., Chief of Pediatric Pulmonology and Critical Care, Elmhurst Hospital Center-Mt. Sinai Services, Mt. Sinai School of Medicine, New York, USA for help in translating the Spanish articles.

Potential conflict of interest

  • None noted.

[top]

Characteristics of studies

Characteristics of Included Studies

Study Methods Participants Interventions Outcomes Notes Allocation concealment
Jonsson 1998 Quasi - randomized (alternate assignment). Concealed randomization- No. Masked intervention-No (placebo not used). Complete follow-up-Yes. Masked outcome-No. 28 intubated preterm infants < 30 weeks gestation culture positive for U. urealyticum. 14 treated, 14 control Treated every other colonized infant with a 10 day course of erythromycin, 40 mg/kg/day, orally (n=10) or intravenously (n=4). Average age at treatment 7 days. Controls received no treatment. Primary outcome: Oxygen dependency at 36 weeks PMA. Secondary outcome: Side effects of erythromycin treatment (GI, hepatic, cardiovascular) C
Lyon 1998 Randomized trial
(sealed envelopes). Concealed randomization -
Yes. Masked intervention-No (placebo not used). Complete follow-up-Yes. Masked outcome-Yes.
75 intubated infants equal to or less than 30 weeks gestation. Of the 75 randomized infants, 9 infants proved to be positive for U. urealyticum by culture and/or PCR. Erythromycin (45 mg/kg/day TID, IV for 7 days) treatment versus no treatment before culture/PCR results were known.
Cytokine levels and differential counts done on bronchoalveolar lavage.
Primary outcome: Inflammatory cytokine response after birth in bronchoalveolar lavage fluid. A

Characteristics of excluded studies

Study Reason for exclusion
Agarwal 2000 No control group.
Alfa 1995 Treatment was not randomized.
Bowman 1998 No control group.
Cordero 1994 Treatment was not randomized.
Cordero 1996 Treatment was not randomized.
Da Silva 1997 Treatment was not randomized.
Galetto Lacour 2001 Treatment was not randomized.
Garland 1996 Treatment was not randomized.
Heggie 1994 Treatment was not randomized.
Izraeli 1991 Treatment was not randomized.
Pacifico 1997 Treatment was not randomized.
Patterson 1998 Treatment was not randomized.
Rudd 1984 No treatment with erythromycin was involved.
Sanchez 1988 No treatment with erythromycin was involved.
Tay 1997 No treatment with erythromycin was involved.

[top]

References to studies

Included studies

Jonsson 1998

{published and unpublished data}

Jonsson B, Rylander M, Faxelius G. Ureaplasma urealyticum, erythromycin and respiratory morbidity in high-risk preterm neonates. Acta Paediatr 1998;87:1079-84.

Lyon 1998

{published and unpublished data}

Lyon AJ, McColm J, Middlemist L, Fergusson S, McIntosh N, Ross PW. Randomised trial of erythromycin on the development of chronic lung disease in preterm infants. Arch Dis Child Fetal Neonatal Ed 1998;78:F10-14.

References to excluded studies

Agarwal 2000

{published data only}

Agarwal P, Rajadurai VS, Pradeepkumar VK, Tan KW. Ureaplasma urealyticum and its association with chronic lung disease in Asian neonates. J Paediatr Child Hlth 2000;36:487-90.

Alfa 1995

{published data only}

Alfa MJ, Embree JE, Degagne P, Olson N, Lertzman J, Macdonald KS, Macdonald NT, Hall PF. Transmission of Ureaplasma urealyticum from mothers to full andpreterm infants. Pediatr Infect Dis J 1995;14:341-5.

Bowman 1998

{published data only}

Bowman ED, Dharmalingam A, Fan WQ, Brown F, Garland SM. Impact of erythromycin on respiratory colonization of Ureaplasma urealyticum and the development of chronic lung disease in extremely low birth weight infants. Pediatr Infect Dis J 1998;17:615-20.

Cordero 1994

{published and unpublished data}

Cordero L, Warner W, Ayers L. Ureaplasma urealyticum tracheal colonization of mechanically ventilated newborn (abstract). Pediatr Res 1994;35:267A.

Cordero 1996

{published data only}

Cordero L, Davis K, Warner D, Ayers L. Ureaplasma urealyticum tracheal colonization and respiratory disease in newborn. Neonatal Intensive Care 1996;9:24-31.

Da Silva 1997

{published data only}

Da Silva O, Gregson D, Hammerberg O. Role of Ureaplasma urealyticum and Chlamydia trachomatis in development of bronchopulmonary dysplasia in very low birth weight infants. Pediatr Infect Dis J 1997;19:364-9.

Galetto Lacour 2001

{published data only}

Galetto Lacour A, Zamora S, Bertrand R, Brighi Perret L, Auckenthaler R, Berner M, Suter S. Colonization by Ureaplasma urealyticum and chronic lungdisease in premature newborn infants under 32 weeks of gestation. Arch Pediatr2001;8:39-46.

Garland 1996

{published data only}

Garland SM, Bowman ED. Role of Ureaplasma urealyticum and Chlamydia trachomatis in lung disease in low birth weight infants. Pathology 1996;28:266-69.

Heggie 1994

{published data only}

Heggie AD, Jacobs MR, Butler VT, Baley JE, Boxerbaum B. Frequency and significanceof isolation of Ureaplasma urealyticum and Mycoplasma hominis from cerebrospinalfluid and tracheal aspirate specimens from low birth weight infants. J Pediatr1994;124:956-61.

Izraeli 1991

{published data only}

Izraeli S, Samra Z, Sirota L, Merlob P, Davidson S. Genital mycoplasmas in preterm infants: prevalence and clinical significance. Eur J Pediatr 1991;150:804-7.

Pacifico 1997

{published data only}

Pacifico L, Panero A, Roggini M, Rossi N, Bucci G, Chiesa C. Ureaplasma urealyticum and pulmonary outcome in a neonatal intensive care population. Pediatr Infect Dis J 1997;16:579-86.

Patterson 1998

{published data only}

Patterson AM, Taciak V, Lovchik J, Fox RE, Campbell AB, Viscardi RM. Ureaplasma urealyticum respiratory tract colonization is associated with an increase in interleukin 1-beta and tumor necrosis factor alpha relative to interleukin 6 in tracheal aspirates of preterm infants. Pediatr Infect Dis J 1998;17:321-8.

Rudd 1984

{published data only}

Rudd PT, Carrington D. A prospective study of chlamydial, mycoplasmal, and viral infections in a neonatal intensive care unit. Arch Dis Child 1984;59:120-5.

Sanchez 1988

{published data only}

Sanchez PJ, Regan JA. Ureaplasma urealyticum colonization and chronic lungdisease in low birth weight infants. Pediatr Infect Dis J 1988;7:542-6.

Tay 1997

{published data only}

Tay ST, Boo NY, Khoo TB, Koay AS, Rohani MY. Prevalence and antibiotic susceptibility of Ureaplasma urealyticum in Malaysian neonates with respiratory distress. Med J Malaysia 1997;52:409-11.

References to studies awaiting assessment

Okura 1998

{published data only}

Okura M, Sato Y, Takeuchi Y, Ono T, Hashimoto T, Nishigaki K. Multicenter study on clinical effects of a small dose of erythromycin on chronic lung disorders of premature infants-single blind comparison of treated and treated groups. A committee for the Study of EM Therapy for Chronic Lung Disease of Premature Infants. Jpn J Antibiot 1998;51 Suppl 1:103-6.

* indicates the primary reference for the study

Other references

Additional references

Blanchard 1993

Blanchard A, Hentschel J, Duffy L, Baldus K, Cassell GH. Detection of Ureaplasmaurealyticum by polymerase chain reaction in the urogenital tract of adults, in amniotic fluid, and in the respiratory tract of newborns. Clin InfectDis 1993;17:S148-53.

Buhrer 2001

Buhrer C, Hoehn T, Hentschel J. Role of erythromycin for treatment of incipientchronic lung disease in preterm infants colonised with Ureaplasma urealyticum.Drugs 2001;61:1893-99.

Cassell 1988a

Cassell GH, Crouse DT, Waites KB, et al. Does Ureaplasma urealyticum cause respiratory disease in newborns? Pediatr Infect Dis J 1988;7:535-41.

Cassell 1988b

Cassell GH, Waites KB, Crouse DT, et al. Association of Ureaplasma urealyticum infection of the lower respiratory tract with chronic lung disease and death in very-low-birth-weight infants. Lancet 1988;2:240-5.

Cassell 1991

Cassell GH, Waites KB, Crouse DT. Perinatal mycoplasmal infections. Clin Perinatol 1991;18:241-62.

Cassell 2001

Cassell GH, Waites KB, Crouse DT. Mycoplasmal infections. In: Remington JS, Klein JO, editor(s). Infectious Diseases of the Fetus and Newborn Infant. 5th edition. Philadelphia: Saunders, 2001:733-67.

Crouse 1993

Crouse DT, Odrezin GT, Cutter GR, Reese JM, Hamrick WB, Waites KB, Cassell GH. Radiographic changes associated with tracheal isolation of Ureaplasma urealyticum from neonates. Clin Infect Dis 1993;17 Suppl 1:S122-30.

Cunliffe 1996

Cunliffe NA, Fergusson S, Davidson F, Lyon A, Ross PW. Comparison of culture with the polymerase chain reaction for detection of Ureaplasma urealyticum in endotracheal aspirates of preterm infants. J Med Microbiol 1996;45:27-30.

Dinsmoor 1989

Dinsmoor MJ, Ramamurthy RS, Gibbs RS. Transmission of genital mycoplasmas from mother to neonate in women with prolonged membrane rupture. Pediatr InfectDis J 1989;8:483-7.

Forgacs 1993

Forgacs P, Kundsin RB, Margles SW, Silverman ML, Perkins RE. A case of Ureaplasma urealyticum septic arthritis in a patient with hypogammaglobulinemia. Clin Infect Dis 1993;16:293-4.

Garland 1987

Garland SM, Murton LJ. Neonatal meningitis caused by Ureaplasma urealyticum. Pediatr Infect Dis J 1987;6:868-70.

Grenabo 1988

Grenabo L, Hedelin H, Pettersson S. Urinary infections caused by Ureaplasma urealyticum: a review. Scand Infect Dis Suppl 1988;53:46-9.

Hannaford 1999

Hannaford K, Todd DA, Jeffery H, John E, Blyth K, Gilbert GL. Role of Ureaplasma urealyticum in lung disease of prematurity. Arch Dis Child Fetal Neonatal Ed 1999;81:F162-7.

Heggie 2001

Heggie AD, Bar-shain D, Boxerbaum B, Fanaroff A, et al. Identification and quantification of ureaplasmas colonizing the respiratory tract and assessment of their role in the development of chronic lung disease in preterm infants. Pediatr Infect Dis J 2001;20:854-9.

Honein 1999

Honein MA, Paulozzi LJ, Himelright IM, Lee B, Cragan JD, Patterson L, CorreaA, Hall S, Erickson JD. Infantile hypertrophic pyloric stenosis after pertussisprophylaxis with erythromycin: a case review and cohort study. Lancet 1999;354:2101-5.

Horowitz 1992

Horowitz S, Landau D, Shinwell ES, Zmora E, Dagan R. Respiratory tract colonization with Ureaplasma urealyticum and bronchopulmonary dysplasia in neonates in southern Israel. Pediatr Infect Dis J 1992;11:847-51.

Iles 1996

Iles R, Lyon A, Ross P, McIntosh N. Infection with Ureaplasma urealyticum and Mycoplasma hominis and the development of chronic lung disease in preterm infants. Acta Paediatr 1996;85:482-4.

Iles 1997

Iles R, Edmunds AT. Assessment of pulmonary function in resolving chronic lung disease of prematurity. Arch Dis Child 1997;76:F113-17.

Kenyon 2002

Kenyon S, Boulvain M, Neilson J. Antibiotics for preterm premature rupture of membranes (Cochrane Review). In: The Cochrane Library, Issue 4, 2002. Oxford:Update Software.

McCormack 1986a

McCormack WM, Rosner B, Alpert S, Evrard JR, Crockett VA, Zinner SH. Vaginal colonization with Mycoplasma hominis and Ureaplasma urealyticum. Sex Transm Dis 1986;13:67-70.

McCormack 1986b

McCormack WM. Ureaplasma urealyticum: ecologic niche and epidemiologic considerations. Pediatr Infect Dis 1986;5:S232-3.

Mohiuddin 1991

Mohiuddin AA, Corren J, Harbeck RJ, Teague JL, Volz M, Gelfand EW. Ureaplasma urealyticum chronic osteomyelitis in a patient with hypogammaglobulinemia. J Allergy Clin Immunol 1991;87:104-7.

Northway 1967

Northway WHJ, Rosan RC, Porter DY. Pulmonary disease following respirator therapy of hyaline-membrane disease. Bronchopulmonary dysplasia. N Engl J Med 1967;276:357-68.

Ollikainen 2001

Ollikainen J, Korppi M, Heiskanen-Kosma T, Heinonen K. Chronic lung disease of the newborn is not associated with Ureaplasma urealyticum. Pediatr Pulmonol 2001;32:303-7.

Payne 1993

Payne NR, Steinberg SS, Ackerman P, Chrenka BA, Sane SM, Anderson KT, FangmanJJ. New prospective studies of the association of Ureaplasma urealyticum colonizationand chronic lung disease. Clin Infect Dis 1993;17 Suppl 1:S117-21.

Perzigian 1998

Perzigian RW, Adams JT, Weiner GM, Dipietro MA, Blythe LK, Pierson CL, Faix RG. Ureaplasma urealyticum and chronic lung disease in very low birth weight infants during the exogenous surfactant era. Pediatr Infect Dis J 1998;17:620-5.

Sanchez 1987

Sanchez PJ, Regan JA. Vertical transmission of Ureaplasma urealyticum infull term infants. Pediatr Infect Dis J 1987;6:825-8.

Sanchez 1988

Sanchez PJ, Regan JA. Ureaplasma urealyticum colonization and chronic lungdisease in low birthweight infants. Pediatr Infect Dis J 1988;7:542-46.

Sanchez 1990

Sanchez PJ, Regan JA. Vertical transmission of Ureaplasma urealyticum frommothers to preterm infants. Pediatr Infect Dis J 1990;9:398-401.

Saxen 1993

Saxen H, Hakkarainen K, Pohjavuori M, Miettinen A. Chronic lung disease of preterm infants in Finland is not associated with Ureaplasma urealyticum colonization. Acta Paediatr 1993;82:198-201.

Shennan 1988

Shennan AT, Dunn MS, Ohlsson A, Lennox K, Hoskins EM. Abnormal pulmonary outcomes in premature infants: prediction from oxygen requirement in the neonatalperiod. Pediatrics 1988;82:527-32.

Smyth 1993

Smyth AR, Shaw NJ Pratt BC, Weindling AM. Ureaplasma urealyticum and chroniclung disease. Eur J Pediatr 1993;152:931-2.

Stancombe 1993

Stancombe BB, Walsh WF, Derdak S, Dixon P, Hensley D. Induction of human neonatal pulmonary fibroblast cytokines by hyperoxia and Ureaplasma urealyticum. Clin Infect Dis 1993;17:S154-7.

Syrogiannopoulos '90

Syrogiannopoulos GA, Kapatais-Zoumbos K, Decavalas GO, et al. Ureaplasma urealyticum colonization of full term infants: perinatal acquisition and persistence during early infancy. Pediatr Infect Dis J 1990;9:236-40.

Tully 1993

Tully JG. Current status of the Mollicute flora of humans. Clin Infect Dis 1993;17 Suppl 1:S2-9.

van Waarde 1997

van Waarde WM, Brus F, Okken A, Kimpen JL. Ureaplasma urealyticum colonization, prematurity and bronchopulmonary dysplasia. Eur Respir J 1997;10:886-90.

Waites 1989

Waites KB, Crouse DT, Philips JB, Canupp KC, Cassell GH. Ureaplasmal pneumonia and sepsis associated with persistent pulmonary hypertension of the newborn. Pediatrics 1989;83:79-85.

Wang 1988

Wang EE, Frayha H, Watts J, et al. Role of Ureaplasma urealyticum and otherpathogens in the development of chronic lung disease of prematurity. PediatrInfect Dis J 1988;7:547-51.

Wang 1995

Wang EE, Ohlsson A, Kellner JD. Association of Ureaplasma urealyticum colonizationwith chronic lung disease of prematurity: results of a metaanalysis. J Pediatr1995;127:640-4.

[top]

Data and analyses

01 Prophylactic erythromycin vs placebo in babies at risk for U. Urealyticum colonization or infection

Comparison or outcome Studies Participants Statistical method Effect size
01.01 CLD 2 75 RR (fixed), 95% CI 1.40 [0.72, 2.70]
01.02 Death 2 75 RR (fixed), 95% CI 0.64 [0.22, 1.83]
01.03 CLD or death 2 75 RR (fixed), 95% CI 1.06 [0.66, 1.69]
01.04 Side effects of erythromycin 0 0 RR (fixed), 95% CI No numeric data

02 Erythromycin vs placebo in U. Urealyticum culture positive babies

Comparison or outcome Studies Participants Statistical method Effect size
02.01 CLD 1 28 RR (fixed), 95% CI 0.90 [0.54, 1.50]
02.02 Death 0 0 RR (fixed), 95% CI No numeric data
02.03 CLD or death 1 28 RR (fixed), 95% CI 0.90 [0.54, 1.50]
02.04 Side effects of erythromycin 0 0 RR (fixed), 95% CI No numeric data

Additional tables

  • None noted.

Amended sections

Cover sheet
Synopsis
Abstract
Background
Objectives
Criteria for considering studies for this review
Search strategy for identification of studies
Methods of the review
Description of studies
Methodological quality of included studies
Results
Discussion
Reviewers' conclusions
References to studies
Other references
Characteristics of Included Studies
Comparisons, data or analyses

Contact details for co-reviewers

Dr Dale L Phelps

Pediatrics and Ophthalmology
University of Rochester
Box 651, Pediatrics
601 Elmwood Ave
Rochester
New York USA
14642
Telephone 1: 716 275 5884
Telephone 2: 716 381 2429
Facsimile: 716 461 3614

E-mail: dale_phelps@urmc.rochester.edu

Dr Gloria S. Pryhuber

Associate Professor
Pediatrics-Neonatology
Strong Children's Research Center, University of Rochester
601 Elmwood Avenue, Division of Neonatology, Box 651
Rochester
NY USA
14642
Telephone 1: 585 275 5884
Facsimile: 585 461 3614

E-mail: gloria_pryhuber@urmc.rochester.edu
Secondary contact person's name: Dale L. Phelps, M.D.

Dr Geoffrey A. Weinberg

Associate Professor
Pediatrics-Infectious Diseases
Strong Children's Research Center, University of Rochester
601 Elmwood Avenue, Box 690, Division of Pediatric Infectious Diseases
Rochester
NY USA
14642
Telephone 1: 585 275 2623

E-mail: geoff_weinberg@urmc.rochester.edu
Secondary contact person's name: Dale L. Phelps, M.D.


This review is published as a Cochrane review in The Cochrane Library, Issue 4, 2003 (see http://www.thecochranelibrary.com External Web Site Policy for information). Cochrane reviews are regularly updated as new evidence emergesand in response to comments and criticisms, and The Cochrane Library shouldbe consulted for the most recent version of the Review.